Download 42 8 Angefügte Publikationen und Manuskripte Schilling, S

8 Angefügte Publikationen und Manuskripte
Schilling, S., Hoffmann, T., Rosche, F., Manhart, S., Wasternack, C., Demuth, H.-U. (2002a).
Heterologous expression and characterization of human glutaminyl cyclase: evidence for a
disulfide bond with importance for catalytic activity. Biochemistry 41, 10849-10857.
Schilling, S., Hoffmann, T., Wermann, M., Heiser, U., Wasternack, C., Demuth, H.-U. (2002b).
Continuous spectrometric assays for glutaminyl cyclase activity. Anal. Biochem. 303, 49-56.
Schilling, S., Manhart, S., Hoffmann, T., Ludwig, H.-H., Wasternack, C., Demuth, H.-U.
(2003a). Substrate specificity of glutaminyl cyclases from plants and animals. Biol. Chem.
384, 1583-1592.
Schilling, S., Niestroj, A. J., Rahfeld, J.-U., Hoffmann, T., Wermann, M., Zunkel, K.,
Wasternack, C., Demuth, H.-U. (2003b). Identification of human glutaminyl cyclase as a
metalloenzyme: Inhibition by imidazole derivatives and heterocyclic chelators. J. Biol. Chem.
278, 49773-49779.
Schilling, S. und Demuth, H.-U. (2004). Continuous assays of glutaminyl cyclase: From
development to application. Spectroscopy, im Druck.
Schilling, S., Hoffmann, T., Manhart, S., Hoffmann, M., Demuth, H.-U. Inhibition of
glutaminyl cyclase prevents the formation of pGlu3-amyloid-β related peptides.
Biochemistry, in Revision.
42
Biochemistry 2002, 41, 10849-10857
10849
Heterologous Expression and Characterization of Human Glutaminyl Cyclase:
Evidence for a Disulfide Bond with Importance for Catalytic Activity
Stephan Schilling,‡ Torsten Hoffmann,‡ Fred Rosche,‡ Susanne Manhart,‡ Claus Wasternack,§ and
Hans-Ulrich Demuth*,‡
Laboratory of Biochemistry, Probiodrug AG, Weinbergweg 22, 06120 Halle/Saale, Germany,
and Leibniz Institute for Plant Biochemistry, P.O. Box 110432, 06120 Halle/Saale, Germany
ReceiVed April 29, 2002; ReVised Manuscript ReceiVed July 10, 2002
ABSTRACT: Glutaminyl cyclase (QC, EC 2.3.2.5) catalyzes the formation of pyroglutamate residues from
glutamine at the N-terminus of peptides and proteins. In the current study, human QC was functionally
expressed in the secretory pathway of Pichia pastoris, yielding milligram quantities after purification
from the supernatant of a 5 L fermentation. Initial characterization studies of the recombinant QC using
MALDI-TOF mass spectrometry revealed correct proteolytic processing and N-glycosylation at both
potential sites with similar 2 kDa extensions. CD spectral analysis indicated a high R-helical content,
which contrasts with plant QC from Carica papaya. The kinetic parameters for conversion of H-GlnTyr-Ala-OH by recombinant human QC were almost identical to those previously reported for purified
bovine pituitary QC. However, the results obtained for conversion of H-Gln-Gln-OH, H-Gln-NH2, and
H-Gln-AMC were found to be contradictory to previous studies on human QC expressed intracellularly
in E. coli. Expression of QC in E. coli showed that approximately 50% of the protein did not contain a
disulfide bond that is present in the entire QC expressed in P. pastoris. Further, the enzyme was consistently
inactivated by treatment with 15 mM DTT, whereas deglycosylation had no effect on enzymatic activity.
Analysis of the fluorescence spectra of the native, reduced, and unfolded human QC point to a
conformational change of the protein upon treatment with DTT. In terms of the different enzymatic
properties, the consequences of QC expression in different environments are discussed.
Besides proteolytic cleavage and C-terminal amidation,
N-terminal formation of 5-oxoproline (pyroglutamate, pGlu)
is a common post-translational event during the biosynthesis
of a number of peptides. Examples of pyroglutamate
containing peptides include the hormones thyrotropin-releasing hormone (TRH) and gastrin, the neuropeptide neurotensin, and the chemokines MCP-1 through MCP-4 (1, 2).
For peptides such as TRH and MCP-2, biological function
has been shown to depend on the 5-oxoproline at their
N-terminus. Loss or modification of this residue leads to a
decrease in biological activity (3, 4). The maturation of these
peptide hormones, taking place in the regulated secretory
pathway (RSP), is well understood, and many of the enzymes
involved in the pro-hormone to hormone conversion have
been identified and characterized (5, 6). The enzyme
glutaminyl cyclase (QC), however, responsible for formation
* Corresponding author. Tel.: 49 345 5559900. Fax: 49 345
5559901. E-mail: [email protected].
‡ Probiodrug AG.
§ Leibniz Institute for Plant Biochemistry.
1
Abbreviations: BMGY, buffered glycerol complex medium;
BMMY, buffered methanol complex medium; DAHC, diammonium
hydrogen citrate; DAP A, dipeptidyl aminopeptidase A; DHAP, 2′,6′dihydroxyacetophenone; DTT, dithiothreitol; GdmCl, guanidinium
chloride; GlcNAc, N-acetyl-D-glucosamine; H-Gln-AMC, L-glutaminyl4-methylcoumarinylamide; H-Gln-βNA, L-glutaminyl-2-naphthylamide;
IMAC, immobilized metal ion affinity chromatography; IPTG, isopropyl
β-D-1-thiogalactopyranoside; Man, D-mannose; MCP, monocyte chemotactic protein; QC, glutaminyl cyclase; TFA, trifluoroacetic acid; TRH,
thyrotropin releasing hormone
of pyroglutamate from glutamine at the N-termini of
hormones, is poorly understood.
First identified in the plant Carica papaya (7), QCs have
been reported from a bovine (8), porcine (9), and human as
well as other mammalian sources (10, 11). Though the QCs
from plants and mammals are similar in size (33 kDa and
38-40 kDa respectively), recent studies have revealed little
or no sequence homology between them (12). A highly
conserved primary structure, however, was reported for QCs
from different mammalian species (11).
Due to the abundancy of QC in papaya latex and to a
simple isolation procedure, the majority of biochemical data
of QC has been collected for the papaya enzyme (13-15).
To date, there has been only one report of QC purification
to homogeneity from a natural mammalian source. On the
basis of considerable effort, 38 µg of homogeneous QC could
be recovered from 2000 bovine pituitaries (16). Subsequently,
cDNAs encoding the bovine and human enzyme, respectively, have been isolated, and enzymological studies with
recombinant human QC expressed intracellulary in Escherichia coli have been reported (10, 11). Aggregation of the
protein during bacterial expression, however, has necessitated
protein refolding under denaturing conditions or expression
of the QC as a fusion protein (N-terminal mannose binding
protein or glutathione S-transferase) in order to recover the
active protein.
These difficulties tempted us to express human QC in
another host system, Pichia pastoris. This methylotrophic
10.1021/bi0260381 CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/09/2002
10850 Biochemistry, Vol. 41, No. 35, 2002
Schilling et al.
Table 1: Oligonucleotides Used in Cloning Procedures for Heterologous Expression of QC in E. coli and P. pastoris
oligonucleotiode
sequence (5′ f 3′), restriction sites (underlined)
restriction enzyme
for cloning
QC-SDMCs
QC-SDMCas
QC4
QC5
pQCyc-1
pQCyc-2
QC-Pic1
QC-Pic2
HPic-K1
HPic-K2
GCCGTGCCATGTGCAATGATGTTG
CAACATCATTGCACATGGCACGGC
ATAGTCGACGCAGGCGGAAGACACCGGC
ATAAAGCTTTTACAAATGAAGATATTCC
ATATAGCATGCGGAGGAGAAGAATTACCACCAG
ATATAAAGCTTACAAATGAAGATATTCCAACAC
ATGCTAGCGCCTGGCCAGAGGAGAAGAAT
ATTCTAGAGTATTACAAATGAAGATATTC
GCTCATCATCATCATCATCATGCTAGCGGTAC
CGCTAGCATGATGATGATGATGATGAGCTGCA
SalI
HindIII
SphI
HindIII
NheI
XbaI
NheI
NheI
yeast shares the advantages of bacterial hosts, such as simple
genetic manipulation, simple culture conditions and rapid
growth, while facilitating post-translational modification in
a manner more similar to that of higher eukaryotes (e.g.,
N-glycosylation, disulfide formation, fatty acylation and
C-terminal methylation). Further, P. pastoris allows highlevel expression of heterologous proteins either intracellulary
or in secreted form (17-19), as well as allowing simple
scale-up production using a fermenter.
Here, we describe the large-scale expression of human QC
in P. pastoris and the subsequent identification of important
enzymatic properties in contrast to those obtained from
human QC expressed in E. coli.
EXPERIMENTAL PROCEDURES
Host strains and media. The E. coli strain JM109 was
applied for all plasmid constructions and propagation. P.
pastoris strain X33 (AOX1, AOX2), used for the expression
of human QC was grown, transformed, and analyzed
according to the manufacturer’s instructions (Invitrogen).
Media for propagation of E. coli, i.e., low salt LB, as well
as the media required for P. pastoris, i.e., buffered glycerol
(BMGY) complex or methanol (BMMY) complex medium,
and the fermentation basal salts medium were prepared
according to the manufacturer’s recommendations.
Isolation of QC cDNA and site-directed mutagenesis. A
full-length cDNA encoding the human QC was isolated from
clone DKFZp566F243, obtained from the resource center
of the human genome project at the Max-Plank-Institute for
Molecular Genetics (Berlin, Germany). Sequencing (Seqlab
GmbH, Göttingen) of the cDNA revealed three single base
exchanges in the open reading frame, one in codon 15
replacing CTG (Leu) by CCG (Pro), a silent exchange in
codon 98 (CTC instead of CTT), and in codon 164 TGT
(Cys) was replaced by TGG (Trp). Site-directed mutagenesis
was carried out to replace Trp with Cys at position 164. All
amplifications were performed according to the instructions
of the supplier of the Pfu polymerase utilized (New England
Biolabs). In brief, the complementary oligonucleotides QCSDMCs and QC-SDMCas (Table 1) were employed for
amplification of two fragments of the cDNA with oligonucleotides QC4 and QC5, which cover the open reading
frame of QC. Subsequently, a second PCR was performed
using the two fragments as template and again Primers QC4
and QC5, yielding the cDNA encoding a Cys in codon 164.
The open reading frame was subcloned into the pPCR-Script
Cam SK(+) vector applying the manufacturer’s recommendations (Stratagene).
Molecular cloning of plasmid Vectors encoding the human
QC. All cloning procedures were done applying standard
molecular biology techniques (20). For expression in yeast,
the vector pPICZRB (Invitrogen) was used that covers a
coding sequence for the S. cereVisiae R-factor prepro-peptide
upstream of a multiple cloning site. To express the QC with
an N-terminal 6×histidine affinity tag, a cassette consisting
of the oligonucleotides HPic-K1 and HPic-K2 (Table 1) was
inserted in frame with the leader sequence using the
restriction sites PstI and KpnI. In addition, with the cassette
a novel restriction site for NheI was introduced. Finally, the
cDNA encoding the mature QC starting with amino acid 33
was amplified by PCR using the primers QC-Pic1 an QCPic2 (Table 1). Subsequently, subcloning into the pPCRScript
Cam SK (+) vector and insertion into the yeast expression
plasmid via the NheI and XbaI restriction sites was performed.
The pQE-31 vector (Qiagen) was used to express the
human QC in E. coli. The cDNA of the mature QC starting
with codon 38 was fused in frame with the plasmid encoded
6×histidine tag. After amplification utilizing the primers
pQCyc-1 and pQCyc-2 (Table 1) and subcloning, the
fragment was inserted into the expression vector employing
the restriction sites of SphI and HindIII. All expression
plasmids were sequenced using either vector- or cDNAspecific primers.
Transformation of P. pastoris and mini-scale expression.
Plasmid DNA was amplified in E. coli JM109 and purified
according to the recommendations of the manufacturer
(Qiagen). In the expression plasmid used, pPICZRB, three
restriction sites are provided for linearization. Since SacI and
BstXI cut within the QC cDNA, PmeI was chosen for
linearization. 20-30 µg plasmid DNA was linearized with
PmeI, precipitated by ethanol, and dissolved in sterile,
deionized water. A 10 µg sample of the DNA was then
applied for transformation of competent P. pastoris cells by
electroporation according to the manufacturer’s instructions
(BioRad). Selection was done on plates containing 150 µg/
mL Zeocin. One transformation using the linearized plasmid
yielded several hundred transformants.
To test the recombinant yeast clones for QC expression,
recombinants were grown for 24 h in 10 mL conical tubes
containing 2 mL BMGY. Afterward, the yeast was centrifuged and resuspended in 2 mL BMMY containing 0.5%
methanol. This concentration was maintained by addition of
methanol every 24-72 h. Subsequently, QC activity in the
supernatant was determined. The presence of the fusion
protein was confirmed by western blot analysis using an
Expression and Characterization of Human QC
antibody directed against the 6×histidine tag (Qiagen).
Clones that displayed the highest QC activity were chosen
for further experiments and fermentation.
Large-scale expression in a fermenter. Expression of the
QC was performed in a 5 L reactor (Biostad B, B. Braun
biotech), essentially as described in the “Pichia fermentation
process guidelines” (Invitrogen). In brief, the cells were
grown in the fermentation basal salts medium supplemented
with trace salts, and with glycerol as the sole carbon source
(pH 5.5). During an initial batch phase for about 24 h and a
subsequent fed-batch phase for about 5 h, cell mass was
accumulated. Once a cell wet weight of 200 g/L was
achieved, induction of QC expression was performed using
methanol applying the three-step feeding profile recommended by invitrogen for an entire fermentation time of
approximately 60 h. Subsequently, cells were removed from
the QC-containing supernatant by centrifugation at 6000g,
4 °C for 15 min. The pH was adjusted to 6.8 by addition of
NaOH, and the resultant turbid solution was centrifuged at
37000g, 4 °C, for 40 min. In cases of continued turbidity,
an additional filtration step was applied using a cellulose
membrane (pore width 0.45 µm).
Purification of 6×histidine tagged QC expressed in P.
pastoris. The His-tagged QC was first purified by immobilized metal ion affinity chromatography (IMAC). In a
typical purification, 1000 mL of culture supernatant was
applied to a Ni2+-loaded Chelating Sepharose FF column
(1.6 × 20 cm, Pharmacia) that was equilibrated with 50 mM
phosphate buffer, pH 6.8, containing 750 mM NaCl, at a
flow rate of 5 mL/min. After washing with 10 column
volumes of equilibration buffer and 5 column volumes of
equilibration buffer containing 5 mM histidine, the bound
protein was eluted by a shift to 50 mM phosphate buffer,
pH 6.8, containing 150 mM NaCl and 100 mM histidine.
The resulting eluate was dialyzed against 20 mM Bis-Tris/
HCl, pH 6.8, at 4 °C overnight. Subsequently, the QC was
further purified by anion exchange chromatography on a
Mono Q6 column (BioRad), equilibrated with dialysis buffer.
The QC-containing fraction was loaded onto the column
using a flow rate of 4 mL/min. The column was then washed
with equilibration buffer containing 100 mM NaCl. The
elution was performed by two gradients, resulting in equilibration buffer containing 240 and 360 mM NaCl in 30 or 5
column volumes, respectively. Fractions of 6 mL were
collected and the purity was analyzed by SDS-PAGE.
Fractions containing homogeneous QC were pooled and
concentrated by ultrafiltration. For long-term storage (-20
°C), glycerol was added to a final concentration of 50%.
Protein was quantified according to the methods of Bradford
or Gill and von Hippel (21, 22).
Expression and purification of QC in E. coli. The construct
encoding the QC was transformed into M15 cells (Qiagen)
and grown on selective LB agar plates at 37 °C. Protein
expression was carried out in LB medium containing 1%
glucose and 1% ethanol at room temperature. When the
culture reached an OD600 of approximately 0.8, expression
was induced with 0.1 mM IPTG overnight. After one cycle
of freezing and thawing, cells were lysed at 4 °C by addition
of 2.5 mg/mL lysozyme in 50 mM phosphate buffer, pH
8.0, containing 300 mM NaCl and 5 mM histidine for
approximately 30 min. The solution was clarified by
centrifugation at 37000g, 4 °C for 30 min, followed by
Biochemistry, Vol. 41, No. 35, 2002 10851
two filtration steps applying cellulose filters for crude and
fine precipitates and an additional filtration using a regenerated cellulose membrane (0.45 µM pore width). The supernatant was applied onto the Ni2+-affinity column according
to the purification of QC expressed in P. pastoris. In contrast
to the aforementioned preparation, one additional washing
step with equilibration buffer containing 15 mM histidine
was implemented. Elution of QC was carried out with 50
mM phosphate buffer containing 150 mM NaCl and 100 mM
histidine. The QC-containing fraction was concentrated by
ultrafiltration and immediately used for further experiments
or stored as described for the QC expressed in P. pastoris.
Synthesis of H-Gln-Tyr-Ala-OH and H-Gln-His-Pro-NH2.
Semi-automated synthesis of the tripeptides was performed
on a 0.5 mmol scale using a peptide synthesizer (Labortec
SP650) and the standard Fmoc-protocol of solid-phase
peptide synthesis. Cycles were modified by using double
couplings (shaking 2 × 24 min) with a 2-fold excess of
Fmoc-Tyr(tBu)-OH or Fmoc-His(Trt)-OH and Fmoc-Gln(Trt)-OH, employing the preloaded Fmoc-Ala-Wang (substitution 1.1 mmol/g) in case of H-Gln-Tyr-Ala-OH or the
Rink Amide MBHA resin (substitution 0.79 mmol/g) in case
of H-Gln-His-Pro-NH2. The Wang resin was preloaded in
our laboratories according to standard procedures. Fmoc
deprotection was carried out by using 20% piperidine in
dimethylformamide (1 × 3 min, 1 × 7 min). The amino
acid couplings were performed by 2-(1H-benzotriazole-1yl)-1,1,3,3,-tetramethyluronium tetrafluoroborate (TBTU) (2
equiv)/diisopropyl ethylamine (4 equiv) activation in dimethylformamide. Cleavage from the resin was accomplished
with a cocktail consisting of 95% trifluoroacetic acid (TFA),
2.5% water, and 2.5% tris-isopropylsilane and yielded 80%
of the crude peptide containing approximately 4% of the
pyroglutaminyl peptide. The crude peptides were precipitated
by cold ether and separated from the pyroglutaminyl peptide
by preparative HPLC with TFA free solvents in order to
avoid further cyclization of the N-terminal glutamine.
Preparative HPLC was performed with a linear gradient of
acetonitrile in water (5-65% acetonitrile over 40 min) on a
250-21 Luna RP18 column. Lyophylization resulted in a
white, fluffy substance. To confirm peptide purity and
identity, analytical HPLC and ESI-MS were employed. CHN
analysis of H-Gln-Tyr-Ala-OH was consistent with the
glutaminyl peptide containing one molecule of TFA and one
molecule of water per molecule of peptide.
Assays for glutaminyl cyclase actiVity. All measurements
were performed with a BioAssay Reader HTS-7000Plus for
microplates (Perkin-Elmer) at 30 °C. QC activity was
evaluated fluorometrically using H-Gln-βNA, essentially as
described (23). The samples consisted of 0.2 mM fluorogenic
substrate and 0.25 U pyroglutamyl aminopeptidase (Unizyme, Hørsholm, Denmark) in 0.2 M Tris/HCl, pH 8.0,
containing 20 mM EDTA (50 mM Tris/HCl, pH 8.0,
containing 5 mM EDTA in case of determination of the
kinetic parameters) and 40-400 ng QC in a final volume of
250 µl. Excitation/emission wavelengths were 320/410 nm.
The assay reactions were initiated by addition of glutaminyl
cyclase. QC activity was determined from a standard curve
of β-naphthylamine under assay conditions. One unit is
defined as the amount of QC catalyzing the formation of 1
µmol pGlu-βNA from H-Gln-βNA per minute under the
described conditions.
10852 Biochemistry, Vol. 41, No. 35, 2002
The kinetic parameters for conversion of H-Gln-Gln-OH,
H-Gln-NH2 (both from Bachem), H-Gln-His-Pro-NH2, and
H-Gln-Tyr-Ala-OH by QC were determined spectrophotometrically using a continuous assay developed on basis of a
previously described method (24) (manuscript in preparation).
Spectral analysis of QC and inVestigation of the disulfide
bond status. CD spectra of human QC in 20 mM phosphate
buffer, pH 6.8, were acquired with a Jasco J-715 spectropolarimeter using quartz cuvettes of 0.1 cm path length. The
mean of 15 scans between 185 and 250 nm was calculated
and corrected by subtraction of the buffer spectra. The
percentage of secondary structure elements was calculated
using the Jasco secondary structure estimation program based
on the method of Yang et al. (25).
For investigation of the structural changes of QC, fluorescence emission spectra were recorded in the wavelength
range between 300 and 400 nm using the luminescence
spectrometer LS 50 B (Perkin-Elmer). The excitation wavelength was 295 nm. Spectra were recorded at 100 nm/min
using slit widths of 3 nm for excitation and 6 nm for
emission.
The disulfide bond status was investigated by SDS-PAGE
using deglycosylated QC. Samples were prepared under
reducing (5% 2-mercaptoethanol) and nonreducing conditions
as described previously (26). Electrophoresis was carried out
in 15% polyacrylamide gels (Mini-Protean 3, BioRad), at
constant voltage (200 V) for 2 h. Additionally, the determination of mercaptans in proteins by a dye reagent, Bis(3carboxy-4-nitrophenyl)disulfide, was performed as described
by Ellman (27).
MALDI-TOF mass spectrometry and deglycosylation assay. Matrix-assisted laser desorption/ionization mass spectrometry was carried out using the Hewlett-Packard G2025
LD-TOF System with a linear time-of-flight analyzer. The
instrument was equipped with a 337 nm nitrogen laser, a
potential acceleration source (5 kV), and a 1.0 m flight tube.
Detector operation was in the positive-ion mode, and signals
were recorded and filtered using LeCroy 9350M digital
storage oscilloscope linked to a personal computer. Samples
(5 µl) were mixed with equal volumes of the matrix solution.
For the matrix solution, we used 2′,6′-dihydroxyacetophenone
(DHAP, Aldrich)/diammonium hydrogen citrate (DAHC,
Fluka), prepared by solving 30 mg DHAP and 44 mg DAHC
in 1 mL acetonitrile/0.1% TFA in water (1/1, v/v). A small
volume (∼1 µl) of the matrix-analyte-mixture was transferred
to a probe tip and immediately evaporated in a vacuum
chamber (Hewlett-Packard G2024A sample prep accessory)
to ensure rapid and homogeneous sample crystallization. For
determination of the molecular mass of QC, bovine serum
albumin was added as an internal standard. The deglycosylation experiments were performed after external calibration
using bovine serum albumin. For the deglycosylation, 8 µg
QC was treated with 4 × 10-5 (0.4 NEB units) units of
endoglycosidase Hf (New England Biolabs) in 0.05 M
sodium citrate, pH 5.5, at room temperature, without
denaturation of the protein. Aliquots were removed at
different times, the deglycosylation process was stopped by
addition of matrix (1:1 v/v), and the respective mass spectra
were evaluated.
Schilling et al.
FIGURE 1: Partial sequences of the cDNA sense strand of the
construct used for expression of QC in P. pastoris and its
corresponding amino acid sequence. The cleavage sites of Kex2
and DAP A are indicated. The proteolytic processing results in a
secreted, 6×histidine tagged QC, starting with alanine 33 of the
native sequence.
RESULTS
Expression of human QC in P. pastoris. As demonstrated
for QC from bovine pituitary and hypothalamus, the enzyme
is directed to the secretory pathway caused by an N-terminal
leader sequence and was colocalized in the secretory granules
with its products of catalysis, e.g., TRH (16, 28). Therefore,
secretory expression should also be employed when QC is
produced in P. pastoris. The N-terminal leader sequences
of proteins of higher eucaryotes often yield only small
quantities of secreted proteins in yeast. To improve the
secretion efficiency in P. pastoris, the mature QC cDNA
was fused to the plasmid-encoded sequence of the R-factor
prepro-peptide from S. cereVisiae, capable of directing
proteins efficiently to the secretory pathway (29). This
sequence is extended by codons for two Glu-Ala and for
one Ala-Ala that are post-translationally cleaved by dipeptidyl aminopeptidase A (DAP A). In addition, the Glu-Ala
repeats are thought to favor the cleavage of the R-factor by
Kex2. A coding sequence of a 6×histidine tag was attached
to the 5′-end of the mature QC cDNA in order to facilitate
the purification. Thus, expression of the construct should
result in a secreted protein with 6 histidine residues at its
N-terminus followed by mature QC starting with amino acid
alanine 33, as illustrated in Figure 1.
To increase the integration frequency of the construct for
overexpression, plasmids were linearized in the sequence of
the AOX1 promoter by an endonuclease treatment using the
PmeI restriction site. Initially, transformants were checked
for integration by PCR, immunodetection, and activity
measurements. Since all clones tested were positive, additional recombinants were only checked by QC activity
evaluation. Finally, three of 100 clones showing the highest
expression level were chosen for scaled-up expression in a
fermenter.
A typical fermentation procedure is documented in Fig. 2
by the time course of the optical density at 600 nm (OD600),
cell wet weight and QC activity appearing in the fermentation
medium. The fermentation consisted of the three stages
glycerol batch, glycerol fed batch, and methanol fed batch.
The glycerol phases were marked by rapid growth of the
yeast cells but lacked any QC activity. The production phase,
starting upon depletion of glycerol and supply of methanol,
was indicated by decreased cell growth and appearance of
first QC activity, indicating that QC expression depended
on activation of the AOX promoter. The enzymatic activity
increased throughout the fermentation.
Expression and Characterization of Human QC
Biochemistry, Vol. 41, No. 35, 2002 10853
Table 2: Purification Scheme of Recombinant Human QC
Following Expression in P. pastoris
purification step
protein QC activity specific activity yield
(mg)
(units)
(units/mg)
(%)
fermenation broth
393.2
immobilized metal ion affinity 41.6
chromatography (IMAC)
ion-exchange chromatography
4.1
(IEC)
362.8
256.8
0.9
6.2
100
71
76.9
18.7
21
FIGURE 2: Time course of OD600, cell wet weight and QC activity
in fermentation of a recombinant strain of P. pastoris expressing
human QC. The fermentation can be subdivided into three stages,
the glycerol batch phase for about 24 h, followed by the glycerol
fed batch phase for about 5 h, and finally by the methanol fed batch
phase.
FIGURE 3: Characterization of human QC using SDS-PAGE. (A)
SDS-PAGE analysis of QC-containing fractions of three different
fermentations after IMAC. Lanes: 1, molecular mass standards
(kDa); 2, 55 h fermentation; 3, after 75 h; 4, after 96 h. (B) SDSPAGE illustrating the purification procedure of human QC.
Lanes: 1, fermentation medium after expression; 2, the QCcontaining fraction after IMAC; 3, purified QC after ion-exchange
chromatography; 4, molecular mass standards. Electrophoresis was
performed in 12% gels using reducing sample preparation as
described elsewhere (41). Proteins were visualized by Coomassie
staining.
FIGURE 4: Deglycosylation of recombinant human QC, monitored
by MALDI-TOF mass spectrometry. 8 µg of QC was treated with
4 × 10-5 units of endoglycosidase Hf in 0.05 M sodium citrate,
pH 5.5, at room temperature. At the times indicated, samples were
removed and diluted with matrix and the mass spectra recorded.
The peak at 42.8 kDa corresponds to the double glycosylated QC.
The peaks at 41.1 and 38.8 kDa represent the less glycosylated
QC forms. The 2-fold charged proteins correspond to the peaks
around 20 kDa.
Upon initial purification of the QC-containing fractions
by affinity chromatography on immobilized nickel ions
(IMAC), there were still apparent impurities by a protein of
about 2 kDa less than the QC-containing band. Since
impurities increased during fermentation, represented by an
increase in this lower band, and appearance of a third band
that became the sole band after 96 h of fermentation, the
QC was purified from cultures grown after 60 h of fermentation. At this stage, only residual impurities were found
(Figure 3A). Further purification was performed by chromatography using a strong anion-exchange resin and a very
broad salt gradient. Despite a surprisingly small yield, 4 mg
of pure QC was ultimately obtained from the 5 L fermentation. The purification procedure is shown in Figure 3B and
Table 2.
Characterization of human QC expressed in P. pastoris.
Western blot analysis following IMAC revealed that QC and
both impurities contained a histidine tag (data not shown).
Since the lowest band of the three species could be separated
by lectin affinity chromatography (data not shown), two of
the three protein species seemed to be less glycosylated forms
of QC. This conclusion was corroborated by MALDI-TOF
mass spectrometry. The recombinant human QC displayed
a relatively broad peak at a molecular mass of 42.8 kDa
(Figure 4A). Upon deglycosylation with endoglycosidase Hf,
two other protein species exhibiting molecular masses of 41.1
kDa and 38.8 kDa were formed consecutively (Figure 4BE). The primary structure of QC reveals two potential
N-glycosylation sites, located at asparagine residues 49 and
296 (10). Thus, in the recombinant QC, both asparagines
were glycosylated with oligosaccharides of about 2 kDa per
residue, suggesting that QC is also expressed as a glycoprotein in mammalian cells.
Using MALDI-TOF mass analysis with bovine serum
albumin as an internal standard, from five independently
recorded mass spectra, a molecular mass of 38,795 ( 19
Da was determined for the quantitatively deglycosylated
enzyme (not shown). This mass corresponds well to the
theoretical value of 38 745 Da, calculated from all amino
acid residues and the two GlcNAc residues remaining after
deglycosylation by endoglycosidase Hf. Therefore, posttranslational modifications other than N-glycosylation are
10854 Biochemistry, Vol. 41, No. 35, 2002
Schilling et al.
FIGURE 5: CD-spectroscopic analysis of the secondary structure
of recombinant human QC. The protein was dissolved in 20 mM
potassium phosphate buffer, pH 6.8. Estimation of the secondary
structure revealed 47% R-helix, 14% β-sheet, and 14% β-turn
content.
Table 3: Kinetic Parameters Determined for Human QC Expressed
in P. pastorisa
QC substrate
Michaelis
constant (µM)
turnover
number (s-1)
H-Gln-Tyr-Ala-OH
H-Gln-His-Pro-NH2
H-Gln-AMC
H-Gln-βNA
H-Gln-NH2
H-Gln-Gln-OH
101 ( 4
90 ( 4
51 ( 3
60 ( 6
409 ( 40
148 ( 5
125 ( 1
83 ( 1
5.4 ( 0.1
18.8 ( 0.7
12.8 ( 0.5
20.7 ( 0.2
a Reactions were carried out at 30 °C in 0.05 M Tris/HCl, pH 8.0,
containing 5 mM EDTA.
improbable during expression, and the N-terminus seems to
be completely processed by Kex2 and DAP A, a frequent
cause of inhomogeneities observed when foreign genes are
expressed in P. pastoris.
Further characterization of recombinant human QC was
performed by applying CD spectroscopy (Figure 5). The
appearance of the spectrum indicates a dominant R-helix
content. The two minima at 208 and 222 nm are characteristic
for proteins that contain a high portion of R-helix in their
overall secondary structure. A calculation of quantities of
R-helix, β-sheet, turn, and random structure revealed an
R-helix content of 47% for the human QC. This amount
contrasts with the 5% content reported for the QC from
papaya latex, indicating completely different folding patterns
for both proteins.
Kinetic parameters were recorded for the recombinant
human QC in order to characterize the catalytic competence
of the enzyme. The values obtained at 30 °C with the
substrates H-Gln-Tyr-Ala-OH, H-Gln-His-Pro-NH2, H-GlnAMC, H-Gln-βNA, H-Gln-NH2, and H-Gln-Gln-OH as
substrates are listed in Table 3. Upon examination at 37 °C,
the kinetic parameters Km and kcat for conversion of H-GlnTyr-Ala-OH shifted to 153 ( 5 µM and 220 ( 2 s-1,
respectively.
Interestingly, the data found are in good agreement with
values determined with QC from papaya latex. For instance,
H-Gln-AMC and H-Gln-βNA were converted with Km values
of 52 ( 5 µM and 43 ( 4 µM and kcat values of 31 and 46
s-1, respectively (23). The kinetic parameters listed in Table
3 are in striking contrast, however, to those found by Song
et al. for human QC, expressed in E. coli (10). In this study,
FIGURE 6: SDS-PAGE of the purification steps of human QC
expressed in E. coli. Lanes: 1, molecular mass standards (kDa); 2,
supernatant after lysis, 3, purified QC after IMAC. Electrophoresis
was performed in 12% gels using reducing sample preparation as
described elsewhere (41).
a Km value of 0.64 mM and a kcat of 50.9 min-1 were obtained
for conversion of H-Gln-Gln-OH. Also, H-Gln-NH2 was
processed differently. The enzyme expressed in P. pastoris
exhibits approximately 3-fold tighter binding and 30-fold
faster turnover compared to recombinant human QC expressed in E. coli (30). Most strikingly, H-Gln-AMC was
not converted at all by the human QC expressed in E. coli
(10), while the recombinant human QC from P. pastoris
cyclized H-Gln-AMC almost comparable to other substrates
(Table 3).
Recombinant expression and characterization of human
QC in E. coli. The remarkably different kinetic properties
of human QC expressed in P. pastoris with that formerly
described for QC expressed in E. coli (10) prompted a more
detailed comparison. The cDNA starting with codon 38 was
cloned into the pQE-31 vector and expressed in the cytosol
with an N-terminal 6×histidine tag. Minimal enzymatic
activity was detected upon expression at 37 °C using LB
broth and induction with 0.1 mM IPTG for 5 h. After
overnight expression at room temperature in LB medium
supplemented with 1% glucose and 1% ethanol, however, a
50-fold increase in QC activity was found. Supernatants of
enzymatically lysed bacteria were clarified by several
centrifugation and filtration steps and the resulting soluble
QC was purified to apparent homogeneity by IMAC in one
step (Figure 6). Usually, about 5 mg QC could be purified
per 2 L culture corresponding to an overall yield of 20%.
Applying H-Gln-βNA as substrate, a Km value of 62 ( 5
µM and a kcat of 7.5 ( 0.3 s-1 were found. Thus, compared
to the human QC expressed in yeast, the enzyme expressed
in E. coli showed an identical Km value, but an approximately
3-fold lower turnover number. A frequent reason for reduced
or abolished activity of proteins expressed heterologously
in E. coli is the lack of post-translational modifications such
as proper disulfide bond formation or N-glycosylation.
Therefore, the presence of disulfide bonds in the recombinant
QCs and the influence of glycosylation on the catalysis were
tested. Interestingly, deglycosylation by endoglycosidase Hf
did not alter the kinetic parameters of QC expressed in yeast
using Gln-βNA as substrate, indicating that the lower activity
was not caused by lack of glycosylation (not shown). QC
contains only two cysteine residues, one in position 139 and
another one in 149. Thus, the disulfide status could be easily
analyzed. In a one-dimensional SDS-PAGE after reducing
and nonreducing sample preparation, disulfide-containing
polypeptides migrate different from their reduced counter-
Expression and Characterization of Human QC
FIGURE 7: Disulfide-status of QC expressed in P. pastoris (A) and
E. coli (B), examined using SDS-PAGE (27). Lanes: 1 and 10,
molecular mass standards (kDa); 2, 3, 8, and 9, sample prepared
under reducing conditions (5% β-mercaptoethanol); 4-7, sample
prepared under nonreducing conditions. Electrophoresis was performed in 15% gels (6 × 8 cm) at constant voltage of 200V for 2
h and protein visualized by Coomassie staining. Due to diffusion
of the reducing agent from lanes 3 and 8 into lanes 4 and 7,
respectively, the band pattern was a mixed type.
parts, indicating the presence of intramolecular disulfide
bonds (26). In the case of QC expressed in yeast and
separated by SDS-PAGE after nonreducing sample preparation (Figure 7A, lanes 4-7), the protein clearly migrated
faster than that after a reducing sample preparation (lanes 2
and 3 and lanes 8 and 9), providing evidence of a disulfide
bond in native human QC. In contrast, two bands of similar
strength were formed after nonreducing sample preparation
in the case of QC expressed in E. coli (Figure 7B, lanes
4-7). This clearly indicates that not all of the QC expressed
in E. coli contained a disulfide bond. Accordingly, color
development using Ellman’s reagent was only detected in
the case of the QC that was expressed in E. coli. The portion
of the protein being free of a disulfide bond was calculated
to be 50% using an absorption coefficient of 13 600 M-1
cm-1 (27).
Effects of disulfide reduction on actiVity and structure of
QC expressed in P. pastoris. Due to the obvious effect of
the disulfide on the active structure of QC, the influence of
reducing agents was examined. In the absence of a reducing
agent, QC activity was constant at pH 6.8 and room
temperature for 2 h (Figure 8). In contrast, QC was readily
inactivated by 15 mM dithiotreitol (DTT) during this time.
The loss of QC activity appeared exponentially and was
independent of the initial amount, suggesting pseudo-firstorder kinetics of inactivation. A shift of the pH from 6.8 to
8.0 accelerated the inactivation 10-fold (not shown), suggesting that the process is favored upon formation of the
thiolate-anion of DTT.
To investigate how the structure of QC is affected by
disulfide bond cleavage, fluorescence spectra of the native,
reduced, and unfolded protein were recorded. Usually, upon
denaturation the emission maximum of proteins exhibits a
tryptophan-mediated shift from a shorter wavelength to about
350 nm which corresponds to the fluorescence maximum of
tryptophan in aqueous solutions (31). The native QC
exhibited its fluorescence maximum at 340 nm (Figure 9A).
Upon complete unfolding of the protein in 6 M GdmCl,
the fluorescence intensity decreased and the emission
maximum shifted to 355 nm, indicating a more hydrophilic
environment of the tryptophan residues compared to the
folded state. Interestingly, reduction of the disulfide with 15
Biochemistry, Vol. 41, No. 35, 2002 10855
FIGURE 8: Inactivation of QC by 15 mM DTT (B). Reactions were
carried out in 0.05 M potassium phosphate, pH 6.8, containing 0.3
M NaCl, and started by addition of 8 µg QC. At times indicated,
samples were withdrawn and diluted 100-fold in 0.05 M Tris/HCl,
pH 8.0, and the residual QC activity was determined using H-GlnβNA as substrate. In absence of a reducing agent, enzyme activity
remained constant (A).
FIGURE 9: Fluorescence emission spectra of QC. (A) Spectra of
0.18 µM enzyme samples recorded in 0.05 M sodium phosphate,
pH 6.8, containing 0.3 M NaCl in absence (a) and presence (b) of
6 M GdmCl after excitation at 295 nm. (B) Respective spectra
recorded in the presence of 15 mM DTT at indicated times after
addition of the reducing agent.
mM DTT at pH 6.8 decreased fluorescence intensity only
slightly, and a change in the fluorescence maximum did not
occur (Figure 9B). No change in fluorescence intensity was
detected in the absence of DTT (not shown). The differences
in the fluorescence spectra of the native, reduced, and
10856 Biochemistry, Vol. 41, No. 35, 2002
denatured QC indicate a small conformational change of the
protein possibly caused by the reduction of the disulfide
bond.
DISCUSSION
Many human proteins cannot be purified from natural
sources in amounts necessary for functional analyses. Heterologous expression is often the only choice to get sufficient
amounts of the protein of interest. Among the various
expression hosts, the methylotrophic yeast P. pastoris has
been used successfully for many human proteins (29). In
the current study, we demonstrate functional expression of
human QC in P. pastoris. During expression, the protein was
directed to the secretory pathway by fusion to the R-leader
of S. cereVisiae and purified from the culture supernatant
by a two-step purification procedure. Although QC activity
was readily detectable when the protein was expressed in
shake flasks, the rate of expression was improved 40-fold
by fermentation. The overall yield of expression, however,
was accompanied by heterogeneities in the glycosylation
pattern of QC during long-term fermentations. Because the
different glycoforms could not be separated efficiently by
lectin-affinity- and ion-exchange chromatography, a fermentation time was chosen in which the altered glycoforms
appeared to be minimal (Figure 3A). For mammalian QCs
glycosylation was first shown for the QC from porcine
pituitary (9, 16). The contribution of post-translational
modifications found (approximately 2-4 kDa) in the case
of the bovine QC (16) corresponds to the extent of glycosylation when the human QC is expressed in yeast. This yield
of glycosylation is also in agreement with the structure of
N-linked oligosaccharides of an invertase expressed as a
heterologous protein in P. pastoris, too. The recombinant
invertase contains more than 85% oligosaccharides in the
size range Man8-14GlcNAc2, thus comparable to highmannose oligosaccharides synthesized by animal cells (19).
In addition, hyperglycosylation that is frequently observed
when heterologous proteins are expressed in S. cereVisiae
is commonly not found with proteins expressed in P. pastoris
(17, 29). Heterogeneities in glycosylation, however, have
been previously reported for other proteins expressed in P.
pastoris, for instance, interleukin-17 (32) and HIV-1 envelope protein (33). Here, the glycosylation pattern of QC
shifted to deglycosylated forms at later stages of growth
(Figure 2). Previously, for S. cereVisiae a deglycosidase
activity was reported that increased in cells reaching the
stationary phase (34). Possibly, the appearance of the
glycoforms of QC could be due to a post-translational
cleavage of the sugar moieties. The human QC expressed in
P. pastoris was deglycosylated quantitatively by only very
low amounts of endoglycosidase Hf (Figure 4). Thus, both
glycosylation sites seem to be easily accessible for a putative
deglycosidase, even in the native, folded structure of QC,
and both are possibly exposed at the protein surface. Human
and bovine QC contain putative glycosylation sites at
positions 49 and at positions 296 and 183, respectively, and
both proteins show an overall sequence identity of 86%. The
glycosylation sites, however, are within less conserved
regions, implying that the protein conformation around these
sites are less important for the catalytic properties of the QC.
This is strengthened by the fact that catalytic parameters were
unaffected by deglycosylation of the enzyme.
Schilling et al.
The nearly identical kinetic parameters of bovine and
recombinant human QC obtained with the peptide H-GlnTyr-Ala-OH also reflect the high degree of homology
between the enzymes. To date, estimates of the Michaelis
constant and the turnover number for a native mammalian
QC can only be achieved for H-Gln-Tyr-Ala-OH from
literature data. On the basis of a molecular weight of 3840 kDa for the purified bovine pituitary protein, a kcat value
of 225-235 s-1 can be calculated for the conversion of
H-Gln-Tyr-Ala-OH at 37 °C (16). The corresponding Michaelis constant of 132 µM was determined in an earlier study
(8). Thus, the kinetic parameters determined using the
recombinant human QC are in excellent agreement with these
earlier results obtained applying the highly homologous QC
from bovine pituitary. This suggests that the proteins have
similar if not identical catalytic competence, despite of
heterologous expression and the presence of an N-terminal
affinity tag in the recombinant human QC.
Remarkable differences in enzymatic activity were found
between human QC expressed in E. coli (10, 11, 30) and
that expressed in yeast, as reported here. At least partially,
these differences might be due to the glutathione Stransferase fused to the N-terminus of QC expressed in E.
coli (30). In the current study, the major portion of the QC
protein expressed in E. coli, however, was inactive, suggesting that the active structure of QC was not formed as
also indicated by the absence of the disulfide bond. This
could be an additional reason for the apparent reduction in
QC activity reported previously. Whether inactive proteins
were formed by improper folding or by a lack of disulfide
formation was not investigated in detail. However, initial
experiments to separate the active and inactive QC forms
by ion exchange chromatography failed. Furthermore, enzymatic activity could not be restored by addition of oxidized
and reduced glutathione, a method often used for refolding
of proteins in order to facilitate the correct formation of the
disulfide bonds (35). Additionally, QC activity could not be
detected when QC cDNA was expressed with the single base
exchange in codon 164, which led to a tryptophan residue
instead of a cysteine in this position. Although these results
could also be interpreted in terms of misfolding, the loss of
the disulfide bond of QC could also be a reason for
inactivation, as indicated by the treatment of QC with DTT.
Reducing cytosolic conditions are known to hinder the
formation of disulfides in proteins (36), and therefore,
translocation of the QC into a less reducing environment
seems to be important for efficient formation of the
enzymatically active structure.
The fluorescence spectra obtained for native, reduced, and
unfolded QC showed that reduction of the disulfide bond
resulted in a relatively small change of the protein conformation, indicated by a reduced fluorescence intensity. The
unchanged fluorescence emission maximum also pointed to
minor conformational differences. The concomitant loss of
the enzymatic activity, however, clearly indicate an important
role of the disulfide bond for the stabilization of the active
protein structure. Furthermore, the fluorescence maximum
at 339 nm of native QC indicates that not all tryptophan
residues are in a hydrophobic environment. In such cases,
the fluorescence maximum can still shift further into the blue
range, as shown for RNase T1 (31) or prolyl oligopeptidase
(37).
Expression and Characterization of Human QC
The QCs from C. papaya and its mammalian counterparts
seem to be very similar with respect to molecular weights,
subunit composition, and catalytic properties (30, 38).
However, there was no sequence homology found between
the enzymes (12), and their folding pattern was assumed to
be different (14). Similarities were found between the
predicted structure of human QC and bacterial zinc-dependent aminopeptidases that contain an R/β-structure (30). In
addition, two different prediction methods (39, 40) used to
calculate the portions of secondary structure from the amino
acid sequence yielded 43% and 52% of R-helix and 15%
and 16% of β-sheet for the recombinant human QC. Thus,
the calculated values from the CD spectroscopy data and
the predicted values are in the same range. Similar characterization experiments performed with QC from papaya latex
revealed that the protein adopts an all-beta structure (14).
Although the β-sheet content cannot be calculated from CD
spectra without uncertainties (31), helical contents are mostly
well reflected. Thus, the mammalian QCs seem to contain a
pronounced R-helical secondary structure, in stark contrast
to papaya QC.
To our knowledge, this is the first mammalian QC
expressed and purified from an eukaryotic host. Due to the
post-translational modifications of QC taking place in P.
pastoris, this expression system has proven to be more
favorable than bacterial expression. This might have also
implications for other disulfide containing proteins that are
expressed heterologously in these organisms. The most
important advantage is provided by the fact that the catalytic
competence of the human QC expressed in yeast is identical
to the highly homologous QC purified from bovine pituitary,
providing evidence that the recombinant protein resembles
native QC very well. Therefore, by this study detailed
enzymological and structural studies of human QC can be
initiated.
ACKNOWLEDGMENT
We thank Dr. B. Gerhartz for searching out clone
DKFZp566F243, the resource center of the human genome
project at the Max-Plank-Institute for Molecular Genetics
(Berlin, Germany) for providing the clone, and Dr. A. Porzel
for the help in recording the CD spectra. The technical
assistance of M. Wermann, J. Bär, I. Schulz and B.
Jaschinsky is gratefully acknowledged. Thanks are due to J.
A. Pospisilik for critical reading of the manuscript.
REFERENCES
1. Awade, A. C., Cleuziat, P., Gonzales, T., and Robert-Baudouy, J.
(1994) Proteins 20, 34-51.
2. Garavelli, J. S. (2000) Nucleic Acids Res. 28, 209-211.
3. Abraham, G. N., and Podell, D. N. (1981) Mol. Cell Biochem. 38
Spec No, 181-190.
4. Van Coillie, E., Proost, P., Van Aelst, I., Struyf, S., Polfliet, M.,
De Meester, I., Harvey, D. J., Van Damme, J., and Opdenakker,
G. (1998) Biochemistry 37, 12672-12680.
5. Steiner, D. F. (1998) Curr. Opin. Chem. Biol. 2, 31-39.
Biochemistry, Vol. 41, No. 35, 2002 10857
6. Nillni, E. A., and Sevarino, K. A. (1999) Endocr. ReV. 20, 599648.
7. Messer, M. (1963) Nature 4874, 1299.
8. Fischer, W. H., and Spiess, J. (1987) Proc. Natl. Acad. Sci. U.S.A.
84, 3628-3632.
9. Busby, W. H. J., Quackenbush, G. E., Humm, J., Youngblood,
W. W., and Kizer, J. S. (1987) J. Biol. Chem. 262, 8532-8536.
10. Song, I., Chuang, C. Z., and Bateman, R. C. J. (1994) J. Mol.
Endocrinol. 13, 77-86.
11. Temple, J. S., Song, I., Burns K. H., and Bateman, R. C. J. (1998)
Korean J. Biol. Sci. 2, 243-248.
12. Dahl, S. W., Slaughter, C., Lauritzen, C., Bateman, R. C., Jr.,
Connerton, I., and Pedersen, J. (2000) Protein Expr. Purif. 20,
27-36.
13. Gololobov, M. Y., Song, I., Wang, W., and Bateman, R. C. J.
(1994) Arch. Biochem. Biophys. 309, 300-307.
14. Oberg, K. A., Ruysschaert, J. M., Azarkan, M., Smolders, N.,
Zerhouni, S., Wintjens, R., Amrani, A., and Looze, Y. (1998) Eur.
J. Biochem. 258, 214-222.
15. Zerhouni, S., Amrani, A., Nijs, M., Smolders, N., Azarkan, M.,
Vincentelli, J., and Looze, Y. (1998) Biochim. Biophys. Acta 1387,
275-290.
16. Pohl, T., Zimmer, M., Mugele, K., and Spiess, J. (1991) Proc.
Natl. Acad. Sci. U.S.A. 88, 10059-10063.
17. Romanos, M. (1995) Curr. Opin. Biotechnol. 6, 527-533.
18. Cregg, J. M., Vedvick, T. S., and Raschke, W. C. (1993)
Biotechnology (N.Y.) 11, 905-910.
19. Grinna, L. S., and Tschopp, J. F. (1989) Yeast 5, 107-115.
20. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular
Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory
Press, Plainview, NY.
21. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254.
22. Gill, S. C., and von Hippel, P. H. (1989) Anal. Biochem. 182,
319-326.
23. Schilling, S., Hoffmann, T., Wermann, M., Heiser, U., Wasternack,
C., and Demuth, H.-U. (2002) Anal. Biochem. 303, 49-56.
24. Bateman, R. C. J. (1989) J. Neurosci. Methods 30, 23-28.
25. Yang, J. T., Wu, C. S., and Martinez, H. M. (1986) Methods
Enzymol. 130, 208-269.
26. Allore, R. J., and Barber, B. H. (1984) Anal. Biochem. 137, 523527.
27. Ellman, G. L. (1958) Arch. Biochem. Biophys. 74, 443-450.
28. Bockers, T. M., Kreutz, M. R., and Pohl, T. (1995) J. Neuroendocrinol. 7, 445-453.
29. Cereghino, J. L., and Cregg, J. M. (2000) FEMS Microbiol. ReV.
24, 45-66.
30. Bateman, R. C., Temple, J. S., Misquitta, S. A., and Booth, R. E.
(2001) Biochemistry 40, 11246-11250.
31. Schmid, F. X. (1989) in Protein Structure: A Practical Approach
(Creighton, T. E., Ed.) pp 251-285, IRL Press, Oxford.
32. Murphy, K. P., Gagne, P., Pazmany, C., and Moody, M. D. (1998)
Protein Expr. Purif. 12, 208-214.
33. Scorer, C. A., Buckholz, R. G., Clare, J. J., and Romanos, M. A.
(1993) Gene 136, 111-119.
34. Suzuki, T., Park, H., Kitajima, K., and Lennarz, W. J. (1998) J.
Biol. Chem. 273, 21526-21530.
35. Rudolph, R., and Lilie, H. (1996) FASEB J. 10, 49-56.
36. Hannig, G., and Makrides, S. C. (1998) Trends Biotechnol. 16,
54-60.
37. Polgar, L. (1995) Biochem. J. 312 (Partt 1), 267-271.
38. Gololobov, M. Y., Wang, W., and Bateman, R. C. J. (1996) Biol.
Chem. Hoppe Seyler 377, 395-398.
39. Geourjon, C., and Deleage, G. (1995) Comput. Appl. Biosci. 11,
681-684.
40. Gibrat, J. F., Garnier, J., and Robson, B. (1987) J. Mol. Biol. 198,
425-443.
41. Laemmli, U. K. (1970) Nature 227, 680-685.
BI0260381
Biol. Chem., Vol. 384, pp. 1583 – 1592, December 2003 · Copyright © by Walter de Gruyter · Berlin · New York
Substrate Specificity of Glutaminyl Cyclases from
Plants and Animals
Stephan Schilling1, Susanne Manhart1,
Torsten Hoffmann1, Hans-Henning Ludwig1,
Claus Wasternack2 and Hans-Ulrich Demuth1,*
Probiodrug AG, Weinbergweg 22,
D-06120 Halle/Saale, Germany
2 Leibniz Institute for Plant Biochemistry,
P.O. Box 110432, D-06120 Halle/Saale, Germany
1
*Corresponding author
Glutaminyl cyclases (QC) catalyze the intramolecular
cyclization of N-terminal glutamine residues of peptides and proteins. For a comparison of the substrate
specificity of human and papaya QC enzymes, a novel continuous assay was established by adapting an
existing discontinuous method. Specificity constants
(kcat/Km) of dipeptides and dipeptide surrogates were
higher for plant QC, whereas the selectivity for oligopeptides was similar for both enzymes. However, only
the specificity constants of mammalian QC were dependent on size and composition of the substrates.
Specificity constants of both enzymes were equally
pH-dependent in the acidic pH-region, revealing a
pKa value identical to the pKa of the substrate, suggesting similarities in the substrate conversion mode.
Accordingly, both QCs converted the L-homoglutaminyl residue in the peptide H-homoGln-Phe-LysArg-Leu-Ala-NH2 and the glutaminyl residues of the
branched peptide H-Gln-Lys(Gln)-Arg-Leu-Ala-NH2
as well as the partially cyclized peptide H-Gln-cyclo(N-Lys-Arg-Pro-Ala-Gly-Phe). In contrast, only
QC from C. papaya was able to cyclize a methylated
glutamine residue, while this compound did not even
inhibit human QC-catalysis, suggesting distinct
substrate recognition pattern. The conversion of the
potential physiological substrates [Gln1]-gastrin,
[Gln1]-neurotensin and [Gln1]-fertilization promoting
peptide indicates that human QC may play a key role
in posttranslational modification of most if not all
pGlu-containing hormones.
Key words: Glutamine cyclotransferase / βHomoglutamine / 5-Oxo-L-proline / Pyroglutamic acid.
(Figure 1). In 1963, the first QC was isolated from the latex of the tropical plant Carica papaya (Messer, 1963).
Twenty four years later, a corresponding enzymatic activity was discovered in mammalian pituitary homogenates
(Busby et al., 1987; Fischer and Spiess, 1987). This enzyme is thought to be responsible for the generation of
the N-termini of peptide hormones and proteins containing pGlu. The conversion of Gln into pGlu by a QC was
demonstrated for the precursors of thyroid hormone-releasing hormone (TRH) and gonadotropin-releasing hormone (GnRH) (Busby et al., 1987; Fischer and Spiess,
1987). Experiments revealed a colocalization of the QC
with its putative catalytic products in bovine pituitary,
thereby supporting a potential function in peptide hormone synthesis. The physiological function of plant QC is
unknown. In case of QC from C. papaya, a role in plant
defense reactions was suggested (El Moussaoui et al.,
2001). Putative QCs from other plants have been identified by sequence comparisons (Dahl et al., 2000), but
physiological functions of these enzymes are yet to be
characterized.
An initial comparison of plant and animal QCs reveals
several similarities. QCs from both sources are monomeric with similar molecular masses of 33 – 40 kDa (Pohl
et al., 1991; Zerhouni et al., 1998). They are glycoproteins
synthesized via the secretory pathway, and the carbohydrates contribute approximately 2 – 4 kDa to their molecular mass (Pohl et al., 1991; Dahl et al., 2000). Furthermore, all QCs show a strict specificity for L-glutamine at
the N-terminal position of the substrates and their kinetic
behavior was found to obey the Michaelis-Menten equation (Consalvo et al., 1988; Pohl et al., 1991; Gololobov
et al., 1996). Interestingly, the primary structures of QC
from C. papaya and that of the highly conserved QC from
mammals did not reveal any sequence homology (Dahl
et al., 2000). In addition, plant QC was described to contain extensive β-sheet structure (Oberg et al., 1998), in
contrast to a recent CD-spectral structure determination
Introduction
Glutaminyl cyclases (QC, EC 2.3.2.5) catalyze the intramolecular cyclization of N-terminal glutamine residues
into pyroglutamic acid (pGlu) with liberation of ammonia
Fig. 1
QC.
The N-Terminal Cyclization of Glutaminyl Peptides by
1584
S. Schilling et al.
of the human enzyme (Schilling et al., 2002a). The structure of the mammalian QCs is remarkably homologous to
the bacterial aminopeptidases (Bateman et al., 2001),
whereas the plant QCs appear to belong to a separate
enzyme family (Dahl et al., 2000). This led to the conclusion that QCs from plants and animals have different evolutionary origins.
Due to the apparent divergence, different properties
and consequently different functions are to be expected
when comparing QC isoforms from plants and animals.
To address this question, we performed a direct comparison of plant and animal QC by analysis of the QC
from C. papaya and the recombinant human QC, including a detailed analysis of catalytic properties. Based on
the previously used discontinuous assay methods, only
limited information concerning substrate specificity and
QC selectivity is currently known. So far, kinetic parameters have only been determined for amino acid derivatives or di- and tripeptides (Bateman, 1989; Song et al.,
1994; Gololobov et al., 1996; Sykes et al., 1999). Gonadotropin-releasing hormone and thyrotropin releasing-hormone have been investigated using HPLC as the
only larger potential substrates of QC (Fischer and
Spiess, 1987).
Here, we report the first comprehensive characterization of longer peptides as QC substrates using a novel
continuous spectrophotometric assay. This method enabled a detailed investigation of the impact of N-terminal
glutamate/glutamine modifications, of the nature of the
C-terminal amino acids and the length of the substrates
on QC-catalysis. Furthermore, kinetic parameters of QCcatalysis of potential physiological substrates were also
determined for the first time.
Results
Spectrophotometric Assay
The discontinuous assay introduced by Bateman (1989)
was modified to a continuous method, applicable for microplates. In the assay, glutamic dehydrogenase and its
substrates α-ketoglutaric acid and NADH/H+ are coupled
to QC activity by liberation of ammonia from the glutamine residue. Accordingly, QC activity is reflected by a
decrease in absorbance at 340 nm caused by the ammonia release and subsequent consumption of NADH/H+
due to formation of glutamate from α-ketoglutaric acid.
The amount of the auxiliary enzyme necessary to obtain a
sufficiently short ‘lag time’, i.e. the time in which a steadystate is reached between the reaction catalyzed by the
investigated enzyme and the reaction catalyzed by the
auxiliary enzyme, was first calculated using a previously
described method (McClure, 1969), an approach that
was already used to develop fluorescence-based continuous assays for QC (Schilling et al., 2002b). Under saturating concentrations of NADH/H+ and α-ketoglutaric
acid (McClure, 1969), 30 U/ml glutamic dehydrogenase
Fig. 2 Progress Curves of the Cyclization of H-Gln-Ala-OH,
Catalyzed by Human QC.
The samples contained 0.3 mM NADH/H+, 14 mM α-ketoglutaric
acid, 30 U/ml glutamic dehydrogenase and 1 mM H-Gln-Ala-OH.
For curves A-D, varying concentrations of QC were applied: A,
10 mU/ml, B, 5 mU/ml, C, 2.5 mU/ml. In case of curve D, QC was
omitted. A linear relationship was obtained between the QC concentration and the observed activity (inset).
was calculated to be a sufficient excess to reach 95% of
the steady-state concentration of ammonia after 20 s, i.e.
a time required for starting the reaction and mixing of the
assay constituents. Thus, linear progress curves should
be observed upon starting data acquisition. Indeed, linear progress curves were detected under these conditions (Figure 2). A linear relationship was observed between the activity and the concentration of QC. The
kinetic parameters obtained for H-Gln-Gln-OH in the
continuous assay (Table 1) were in good agreement with
those obtained with the discontinuous method (Km =
175 ± 18 µM, kcat = 21.3 ± 0.6 s– 1). In addition, the kinetic
parameters for conversion of the substrates H-Gln-AlaOH and H-Gln-NH2 by the papaya QC (Table 1) correspond well to those determined with a direct method at
pH 8.8 and 37 °C (Gololobov et al., 1996).
pH Dependence
The pH dependence of catalysis of human and papaya
QC was investigated under first-order rate law conditions, thus reflecting the impact of pH on the specificity
constant kcat/Km. For this purpose, pyroglutamyl aminopeptidase was used as auxiliary enzyme with Gln-βNA as
substrate in the coupled assay. Pyroglutamyl aminopeptidase was shown to be active and stable between
pH 5.5 – 8.5 (Tsuru et al., 1978), hence, this assay allowed
QC analysis in this pH-range. The rate profiles obtained
fit to classical bell-shaped curves (Figure 3). For human
QC a very narrow pH dependence with an optimum at
about pH 7.8 – 8.0 was observed (Figure 3A), and the rate
Substrate Specificity of Glutaminyl Cyclase
Table 1
1585
Kinetic Parameters of the Conversion of Di-, Tri- and Tetrapeptides by Human and Papaya QC.
Substrate
H-Gln-OH
H-Gln-AMC
H-Gln-βNA
H-Gln-OtBu
H-Gln-NH2
H-Gln-Gly-OH
H-Gln-Ala-OH
H-Gln-Gln-OH
H-Gln-Glu-OH
H-Gln-Val-OH
H-Gln-Tyr-OH
H-Gln-Glu-Tyr-NH2
H-Gln-Gly-Pro-OH
H-Gln-Tyr-Ala-OH
H-Gln-Phe-Ala-NH2
H-Gln-Trp-Ala-NH2
H-Gln-Arg-Gly-Ile-NH2
H-Gln-Asn-Gly-Ile-NH2
H-Gln-Ser-Tyr-Phe-NH2
H-Gln-Arg-Tyr-Phe-NH2
H-Gln-Pro-Tyr-Phe-NH2
H-Gln-His-Tyr-Phe-NH2
H-Gln-Gln-Tyr-Phe-NH2
H-Gln-Glu-Tyr-Phe-NH2
H-Gln-Glu-Ala-Ala-NH2
H-Gln-Glu-Tyr-Ala-NH2
H-Gln-Glu-Ala-Phe-NH2
H-Gln-Glu-Asp-Leu-NH2
H-Gln-Lys-Arg-Leu-NH2
Human QC
Papaya QC
Km
(µM)
kcat
(s– 1)
Ki a
(mM)
kcat/Km
(mM– 1 s– 1)
n.r.
54 ± 2
70 ± 3
1235 ± 74
409 ± 40
247 ± 10
232 ± 5
148 ± 5
359 ± 10
196 ± 5
211 ± 5
79 ± 2
130 ± 5
101 ± 4
69 ± 3
50 ± 2
143 ± 4
172 ± 5
55 ± 3
55 ± 2
1889 ± 152
68 ± 3
41 ± 2
47 ± 4
77 ± 4
69 ± 2
39 ± 3
55 ± 2
54 ± 3
n.r.
5.3 ± 0.1
20.6 ± 0.5
6.7 ± 0.2
12.8 ± 0.5
13.2 ± 0.2
57.2 ± 0.4
20.7 ± 0.2
24.7 ± 0.2
17.2 ± 0.1
94 ± 1
45.1 ± 0.4
25.3 ± 0.2
125 ± 1
109 ± 1
47.0 ± 0.7
33.5 ± 0.4
56.6 ± 0.5
52.8 ± 0.8
29.6 ± 0.3
31.7 ± 1.2
55.4 ± 0.7
41.4 ± 0.4
46 ± 1
46 ± 1
42.1 ± 0.4
39 ± 1
45.8 ± 0.5
33.4 ± 0.5
n.r.
n.r.
n.d.
98 ± 2
1.21 ± 0.07 294 ± 6
n.i.
5.4 ± 0.2
n.i.
31 ± 2
n.i.
53 ± 1
n.i.
247 ± 4
n.i.
140 ± 2
n.i.
58 ± 1
n.i.
88 ± 2
n.i.
446 ± 6
n.i.
524 ± 8
n.i.
195 ± 7
n.i.
930 ± 27
n.i.
1811 ± 64
n.i.
940 ± 24
n.i.
234 ± 4
n.i.
329 ± 7
n.i.
960 ± 38
n.i.
538 ± 14
n.i.
17 ± 1
n.i.
815 ± 26
n.i.
1010 ± 40
n.i.
979 ± 62
n.i.
597 ± 18
n.i.
610 ± 12
n.i.
1000 ± 51
n.i.
833 ± 21
n.i.
619 ± 25
Km
(µM)
kcat
(s– 1)
Ki a
(mM)
kcat/Km
(mM– 1 s– 1)
n.d.
42 ± 1
38 ± 3
223 ± 9
433 ± 13
641 ± 20
158 ± 8
44 ± 3
106 ± 5
n.d.
n.d.
103 ± 4
333 ± 15
63 ± 3
111 ± 5
78 ± 5
123 ± 10
153 ± 9
135 ± 6
124 ± 6
149 ± 14
92 ± 7
45 ± 2
100 ± 4
102 ± 4
113 ± 5
81 ± 3
107 ± 6
118 ± 6
n.d.
39.4 ± 0.4
51.4 ± 1.4
49.4 ± 0.6
44.8 ± 0.4
45.8 ± 0.4
69.8 ± 1.0
43.2 ± 0.7
50.3 ± 0.6
n.d.
n.d.
53.6 ± 0.7
41.7 ± 0.5
104.0 ± 1.0
132.1 ± 0.6
151.8 ± 2.6
49.2 ± 1.7
51.4 ± 0.9
64.9 ± 1.0
48.9 ± 0.7
18.8 ± 0.6
75.9 ± 1.4
52.9 ± 0.7
54.6 ± 0.6
53.7 ± 0.6
44.7 ± 0.5
48.5 ± 0.45
58.5 ± 0.4
48.2 ± 0.8
n.d.
n.d.
1.20 ± 0.08
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.d.
n.d.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
0.23 ± 0.1b
938 ± 13
1353 ± 70
222 ± 6
103 ± 2
71 ± 2
442 ± 16
982 ± 51
475 ± 17
n.d.
n.d.
520 ± 13
125 ± 4
1650 ± 63
1190 ± 48
1946 ± 91
400 ± 19
336 ± 14
481 ± 14
394 ± 13
126 ± 8
825 ± 48
1176 ± 37
546 ± 16
526 ± 15
396 ± 13
599 ± 17
547 ± 27
408 ± 14
Reactions were carried out in Tris/HCl, pH 8.,0 containing 5 mM EDTA or in 0.05 M Tricine/NaOH, pH 8.0, 5 mM EDTA, respectively, at
30 °C (n.r., not reactive; n.d., not determined; n.i., no inhibition).
a Substrate inhibition; b determined under first-order rate law conditions, i.e. at [S]<<K .
m
decreased at more basic pH. Also papaya QC exhibited
optimal activity near pH 8.0. However, no drop of kcat/Km
occurred up to a pH of 8.5 (Figure 3B). Evaluation of rate
profiles revealed that at 23 °C, pKa values were
7.17 ± 0.02 and 7.15 ± 0.02 for human QC and papaya
QC, respectively. These pKa values are identical to the
pKa of the substrate H-Gln-βNA (7.16 ± 0.01). The respecitve pKa values obtained by titration and kinetic
analysis of the pH dependence shifted toward 6.97 ± 0.01
and 7.03 ± 0.02 at 30 °C. Thus, substrate binding and/or
conversion by both QC forms seems to be dependent on
an unprotonated α-amino group.
The shape of the pH-profile of human QC in the basic
pH-range was obviously due to dissociation of a group
with a pKa of approximately 8.5. For papaya QC, the limited stability of the auxiliary enzyme resulted in collection
of insufficient data under basic pH conditions. Consequently, a reliable determination of a potential second
pKa value could not be achieved. This is also reflected by
the fact that fitting the data to a single dissociation model gave a very similar result (pKa 7.13 ± 0.03) compared to
a double dissociation model, indicating that both pKa values are fairly separated.
pH Stability
The stability of the QCs was analyzed by preincubating
both enzymes at 30 °C for 30 min at different pH values
between pH 4 and pH 10, followed by activity tests under standard conditions (Figure 4). The QC from papaya
latex was found to be stable in the pH-range studied
(Zerhouni et al., 1998). In contrast, the human QC was
stable only between pH 7 and 8.5, but exhibited a remarkable instability above pH 8.5 and below pH 6.
Thus, pH 8 seemed to be optimal for activity and stability of plant and human QC. Consequently, this pH value
was used to compare the substrate specificity of both
QCs.
1586
S. Schilling et al.
Substrate Specificity
Fig. 3 pH Dependence of Human (A) and Papaya (B) QC.
Determinations were carried out under first-order rate conditions
using Gln-βNA as substrate. In case of human QC, a buffer system providing a constant ionic strength according to Ellis and
Morrison (1982) was used, consisting of 25 mM MES, 25 mM
acetic acid and 50 mM Tris. Due to a slightly inhibiting effect of
Tris, papaya QC was investigated using a 50 mM MOPS buffer.
The ionic strength was adjusted to 0.05 M by addition of NaCl.
The rate profiles were evaluated by fitting the data according to
equations that account for two dissociating groups. In case of
papaya QC, a pKa value of 7.13 ± 0.03 was obtained by fitting of
the data according to a single dissociation model.
Fig. 4 Effect of pH on the Stability of the QC from Papaya Latex and Human QC.
An enzyme stock solution was diluted 20-fold in 0.1 M buffer of
various pH values (pH 4 – 7, sodium citrate; pH 7 – 10, sodium
phosphate). Enzyme solutions were incubated at 30 °C for
30 min, and subsequently enzymatic activity was analyzed according to the standard protocol.
Di-, Tripeptides and Dipeptide Surrogates The substrate specificity of both QCs was analyzed with 30 potential substrates (Table 1). Nearly all of the short peptides
were more efficiently converted by the papaya QC compared to the human QC. This is particularly obvious in
case of the conversion of L-glutamine. Whereas the plant
enzyme was able to react with L-glutamine, no reactivity
of human QC was detected. Both enzymes were highly
active toward substrates carrying large hydrophobic
residues in the penultimate position of peptides, such as
H-Gln-Tyr-Ala-OH, H-Gln-Phe-Ala-NH2 and H-Gln-TrpAla-NH2, or H-Gln-AMC and H-Gln-βNA, as compared to
other tripeptides or dipeptide substrates. Plant as well as
human QC were remarkably inhibited by H-Gln-βNA at
high concentrations, revealing almost identical Ki values
of about 1.2 mM (Table 1). Minor substrate inhibition was
detected for H-Gln-AMC. Hence, low solubility of the substance hampered determination of the Ki value. A striking
difference in specificity of the plant and the animal QC
was observed for H-Gln-OtBu. Whereas the ester was
converted by the papaya QC with similar specificity compared to dipeptide substrates, its turnover was more than
one order of magnitude slower by human QC.
Oligopeptides In addition to several dipeptides and
tripeptides, a number of putative oligopeptide substrates
was tested with papaya and human QC (Table 1). Interestingly, the overall difference in the specificities between
human and plant QC was not as large for tetrapeptides as
was observed for dipeptide and tripeptide substrates.
This indicates that the amino acids in the 3rd and 4th position still affect the kinetic behavior especially of human
QC, leading to similar specificity constants. However,
H-Gln-Pro-Tyr-Phe-NH2, a tetrapeptide with proline in the
penultimate position, yielded noticeably reduced kcat/Km
values, providing an exception (Table 1). The reduction in
the specificity was more pronounced for human QC,
leading to an approximately 8-fold difference in the
kcat/Km value as compared to papaya QC.
Slightly reduced specificity constants of human QC
were also observed for the conversion of substrates with
a positively charged amino acid on the C-terminal side of
glutamine, such as H-Gln-Arg-Tyr-Phe-NH2, H-Gln-ArgGly-Ile-NH2 and H-Gln-Lys-Arg-Leu-NH2, as compared
to other tetrapeptides. Apparently, the reduced specificity was mainly due to a smaller turnover number. This effect was not observed for the plant enzyme.
In contrast to the selectivity of papaya QC for dipeptides, human QC was more selective for some tetrapeptides. The highest selectivity of human QC was recorded
for peptides containing bulky hydrophobic residues in
the 3rd and 4th amino acid position, which indicate hydrophobic interactions with the enzyme. Comparing the
kinetic parameters for the respective peptides, the altered specificity seems to be mainly due to lower Km values, since the turnover numbers for conversion of the
Substrate Specificity of Glutaminyl Cyclase
Table 2
1587
Dependence of the Kinetic Parameters on the Substrate Size.
Substrate
Human QC
H-Gln-Ala -NH2
H-Gln-Ala-Ala-NH2
H-Gln-Ala-Ala-Ala-Ala-NH2
H-Gln-Ala-Ala-Ser-Ala-Ala-NH2
Papaya QC
Km
(µM)
kcat
(s– 1)
kcat/Km
(mM– 1 s– 1)
Km
(µM)
kcat
(s– 1)
kcat/Km
(mM– 1 s– 1)
155 ± 9
87 ± 3
65 ± 3
79 ± 6
40.1 ± 0.9
76.3 ± 0.7
60.5 ± 0.7
55.3 ± 1.6
259 ± 9
877 ± 22
1174 ± 43
700 ± 33
212 ± 21
164 ± 6
197 ± 8
216 ± 6
62.8 ± 3.0
83.2 ± 1.0
74.6 ± 1.0
78.5 ± 1.0
296 ± 15
507 ± 12
379 ± 10
363 ± 5
Reactions were performed under identical conditions as described in Table 1.
Table 3
Dependence of the Catalytic Parameters kcat and Km on the Ionic Strength for Substrates of Varying Length and Charge.
Substrate
Km
(µM)
kcat
(s– 1)
Papaya QC
0.05 M Tricine-NaOH, pH 8.0
H-Gln-NH2
H-Gln-βNA
H-Gln-Ala-OH
H-Gln-Glu-OH
H-Gln-Trp-Ala-NH2
H-Gln-Arg-Gly-Ile-NH2
H-Gln-Lys-Arg-Leu-NH2
H-Gln-Glu-Asp-Leu-NH2
434 ± 15
36 ± 2
137 ± 7
98 ± 5
79 ± 5
106 ± 8
102 ± 7
109 ± 5
Human QC
0.05 M Tris-HCl, pH 8.0
H-Gln-NH2
H-Gln-βNA
H-Gln-Ala-OH
H-Gln-Glu-OH
H-Gln-Trp-Ala-NH2
H-Gln-Arg-Gly-Ile-NH2
H-Gln-Lys-Arg-Leu-NH2
H-Gln-Glu-Asp-Leu-NH2
442 ± 30
76 ± 4
269 ± 7
373 ± 15
54 ± 3
166 ± 13
51 ± 3
60 ± 2
43.4 ± 0.4
48.8 ± 1.0
69.7 ±.9
45.0 ± 0.5
138 ± 3
52.9 ± 1.2
50 ± 1
52.4 ± 0.7
12.8 ± 0.3
21.7 ± 0.5
54.4 ± 0.5
21.4 ± 0.3
50.8 ± 0.6
31 ± 1
29.4 ± 0.5
46.6 ± 0.5
kcat/Km
(mM– 1s– 1)
100 ± 3
1356 ± 50
509 ± 19
459 ± 18
1747 ± 73
499 ± 26
493 ± 22
481 ± 16
Ki a
(mM)
Km
(µM)
kcat
(s– 1)
kcat/Km
(mM– 1s– 1)
Ki a
(mM)
0.05 M Tricine-NaOH, pH 8.0, 0.5 M KCl
n.i.
1.14 ± 0.05
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
446 ± 10
32 ± 2
143 ± 5
94 ± 3
72 ± 4
65 ± 5
53 ± 2
94 ± 3
45.2 ± 0.3
47.2 ± 0.8
68.1 ± 0.6
44.4 ± 0.3
133 ± 3
48.4 ± 1.0
58.1 ± 0.7
53.6 ± 0.5
101 ± 2
1475 ± 70
480 ± 12
472 ± 12
1847 ± 61
745 ± 42
1096 ± 28
570 ± 13
n.i.
1.33 ± 0.07
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
0.05 M Tris-HCl, pH 8.0, 0.5 M KCl
29 ± 1
285 ± 8
202 ± 3
57 ± 2
941 ± 41
187 ± 9
577 ± 24
777 ± 18
n.i.
1.39 ± 0.08
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
401 ± 14
63 ± 3
357 ± 12
607 ± 36
56 ± 2
91 ± 5
34 ± 1
61 ± 2
12.2 ± 0.1
20.0 ± 0.4
47.6 ± 0.6
18.9 ± 0.5
50.0 ± 0.4
29.8 ± 0.5
31.6 ± 0.3
45.6 ± 0.5
30 ± 1
318 ± 9
133 ± 3
31 ± 1
893 ± 25
327 ± 12
929 ± 19
748 ± 16
n.i.
0.97 ± 0.04
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
Reactions were carried out at 30 °C.
a Substrate inhibition.
peptides were found to be similar. Obviously, the higher
selectivity of human QC is due to stronger binding of the
more hydrophobic substrates to the enzyme.
An increasing kcat/Km-ratio was also found for peptides
of varying length consisting of the N-terminal glutamine
residue and alanine residues as substrates of human QC
(Table 2). While human QC was more selective for substrates of a length up to a pentapeptide, there was no
such a trend with the papaya QC. Human QC was less
active toward a serine-containing peptide, indicating that
also the nature of the substrate side chains of the amino
acids close to the reaction center is of importance.
Influence of Ionic Strength on Catalysis To analyze
the influence of ionic strength on the substrate specificity, the kinetic parameters for cyclization of several substrates were determined in presence and absence of 0.5
M KCl (Table 3). Interestingly, the specificity of both QCs
for substrates with uncharged backbone did not change
significantly by salt addition. For substrates such as HGln-Ala-OH and H-Gln-Glu-OH, however, addition of KCl
decreased specificity in case of the human enzyme. As
indicated by the kinetic parameters, this effect was due
to a higher Km and a slightly reduced kcat value. The papaya QC did not show altered kinetic parameters upon
salt addition. The effect did not seem to be due to the
negatively charged substrate as such, since similar kinetic parameters were found for the negatively charged
peptide H-Gln-Glu-Asp-Leu-NH2. With the positively
charged substrates H-Gln-Arg-Gly-Ile-NH2 and H-GlnLys-Arg-Leu-NH2 addition of salt revealed a positive effect on catalysis for both QCs, as indicated by a lower Km
value and a slightly higher turnover number.
Physiological Substrates Physiological substrates of
papaya QC are unknown. For human QC the compounds
1588
S. Schilling et al.
tested here can be regarded as putative substrates. In
earlier studies, conversion of [Gln1]-thyrotropin releasinghormone ([Gln1]-TRH) and [Gln1]-gonadotropin releasinghormone ([Gln1]-GnRH) was shown for QC from bovine
and porcine pituitary glands (Busby et al., 1987; Fischer
and Spiess, 1987). In addition to the previously studied
hypophysiotropic hormones, the following three potential
physiological substrates of human QC were synthesized
and tested: [Gln1]-gastrin, [Gln1]-neurotensin, and [Gln1]fertilization promoting peptide ([Gln1]-FPP) (Table 3). The
glutaminyl peptides were converted to the respective pyroglutamyl peptides with increasing specificity in order of
their size, i.e. the largest peptide [Gln1]-gastrin with
17 amino acids followed by [Gln1]-neurotensin, [Gln1]GnRH, [Gln1]-TRH and [Gln1]-FPP.
Surprisingly, also the plant QC converted longer substrates with higher efficacy. Possibly there are secondary
binding interactions between the substrate and the enzyme distant from the active site that may influence catalysis.
Peptides Comprising Modified Amino Acids The
specificity of the QCs was further analyzed with peptides
that contain either a modified N-terminal glutaminyl
residue or a modified penultimate amino acid. The conversion of these peptides was investigated qualitatively
by MALDI-TOF mass spectrometry. Due to the cyclization of the glutaminyl residue and its analog, respectively,
a mass difference of the substrate and the product of
catalysis was detected. If ammonia was liberated
equimolarly, the conversion was also analyzed quantitatively using the spectrophotometric assay.
(1) H-Gln-Lys(Gln)-Arg-Leu-Ala-NH2 This branched
peptide contains two glutaminyl residues bound to a lysyl
residue via a peptide- and partial isopeptide bond. Human QC (Figure 5) and papaya QC (not shown) converted
this compound apparently in an identical manner. Both
glutaminyl residues were cyclized into pyroglutamic acid,
and the consistent substrate conversion (Figure 5) indicate the lack of any preference for one residue. Thus, the
specificity of the QCs for the differently bound glutaminyl
residues seems not to differ fundamentally.
(2) H-Gln(NMe)-Phe-Lys-Ala-Glu-NH2 The methylated glutaminyl residue was only converted into a pyroglutamyl residue by papaya QC (not shown). An inhibition of
human QC by the peptide was not detected, proving that
the methylated residue is not recognized by human QC.
(3) H-Glu(OMe)-NA and H-Glu-NA Neither of these
compounds were converted by papaya or human QC under the applied conditions. The O-methylated glutamic
acid residue, however, showed a remarkable instability in
both Tris and Tricine buffers leading to non-enzymatic cyclization, probably due to an increased polarity of the γester. Furthermore, catalysis of both QC forms was not
inhibited by the longer peptides H-Glu(OMe)-Phe-Lys-
Fig. 5 Formation of pGlu-Lys(pGlu)-Arg-Leu-Ala-NH2 from HGln-Lys(Gln)-Arg-Leu-Ala-NH2 through Catalysis by Human QC.
Substrate conversion was followed by monitoring by a time-dependent change in the m/z ratio due to the liberation of ammonia. The sample composition was 0.5 mM substrate, 38 nM QC in
40 mM Tris/HCl, pH 7.7. At the times indicated, samples were removed, mixed with matrix solution (1:1 v/v) and the mass spectra recorded. A very similar dependence was observed in case of
papaya QC. There was no substrate conversion in samples without QC (not shown).
Arg-Leu-Ala-NH2 or H-Glu-Phe-Lys-Arg-Leu-Ala-NH2 up
to 4 mM, demonstrating that glutamic acid or derivatives
are not recognized by both QC forms.
(4) H-Gln-cyclo(N-Lys-Arg-Pro-Ala-Gly-Phe) The
conversion of H-Gln-cyclo(Nε-Lys-Arg-Pro-Ala-GlyPhe), which contains an intramolecular partial isopeptide
bond was analyzed quantitatively, revealing Km values of
240 ± 14 µM and 133 ± 5 µM for human and papaya QC,
respectively. Due to the higher turnover number of conversion by the papaya QC (49.4 ± 0.6 s– 1) compared to
the human QC (22.8 ± 0.6 s– 1), the plant enzyme exhibits
with 372 ± 9 mM– 1s– 1 an approximately 4-fold higher
kcat/Km value than the human QC. Thus, the specificity
constant is only slightly lower compared to substrates of
similar size for papaya QC. The kcat/Km value for human
QC was 95 ± 3 mM– 1s– 1, being about one order of magnitude lower than with substrates of similar size (Table 1).
Possibly, the N-terminal glutaminyl residue is less accessible by the enzyme active site due to steric hindrance of the bulky ring and this may have a stronger effect on catalysis by human QC than in case of the plant
enzyme.
(5) H-homoGln-Phe-Lys-Arg-Leu-Ala-NH2 The Nterminal βhomoglutaminyl residue was converted into a
five-membered lactam ring by catalysis of the human and
the papaya QC. The concomitant liberation of ammonia
was analyzed spectrophotometrically and by MALDI-TOF
mass spectrometry as described before. There was no
liberation of ammonia detected when QC was omitted or
boiled, proving enzymatic catalysis of the cyclization. Interestingly, the QCs from C. papaya (Km = 3.1 ± 0.3 mM,
Substrate Specificity of Glutaminyl Cyclase
Table 4
1589
Kinetic Parameters for Cyclization of Several Putative Physiological Substrates of Human QC.
Substrate
[Gln1]-Gastrin
[Gln1]-Neurotensin
[Gln1]-FPP
[Gln1]-TRH
[Gln1]-GnRH
Human QC
Papaya QC
Km
(µM)
kcat
(s– 1)
kcat/Km
(mM– 1 s– 1)
Km
(µM)
kcat
(s– 1)
kcat/Km
(mM– 1 s– 1)
31 ± 1
37 ± 1
87 ± 2
90 ± 4
53 ± 3
54.1 ± 0.6
48.8 ± 0.4
69.6 ± 0.3
82.8 ± 1.2
69.2 ± 1.1
1745 ± 36
1318 ± 24
800 ± 14
920 ± 27
1305 ± 53
33 ± 2
23 ± 2
151 ± 8
n.d.
54 ± 3
31.6 ± 0.9
37.7 ± 0.6
37.7 ± 0.5
n.d.
72.4 ± 1.0
958 ± 40
1639 ± 116
250 ± 10
n.d.
1337 ± 65
Reactions were carried out as described in Table 1 (n.d., not determined; TRH, thyrotropin releasing-hormone; FPP,
fertilization promoting peptide; GnRH, gonadotropin releasing-hormone).
kcat = 4.0 ± 0.4 s– 1) and human (Km = 2.5 ± 0.2 mM, kcat =
3.5 ± 0.1 s– 1) catalyze the conversion of this peptide
with almost identical kcat/Km values of 1.4 ± 0.1 and
1.3 ± 0.1 mM– 1s– 1, respectively. Thus, the cyclization of the
βhomoglutaminyl residue is catalyzed with an approximately 1000-fold reduced efficiency compared to peptides of similar size with N-terminal glutaminyl residue.
This result suggests that the constitution of the α-carbon
of the substrate is important for substrate recognition by
both QC forms, but not essential. Possibly, the essential
requirement for being a substrate is a γ-amide group and
an unprotonated N-terminal amino group in a distance
and angle suitable for cyclization, a requirement that is
fulfilled by both, N-terminal glutaminyl and βhomoglutaminyl residues.
Discussion
Based on the elucidation of the primary structures of the
enzyme from papaya latex, bovine and human pituitary,
several different properties were detected for QCs of
plant and animal origin (Pohl et al., 1991; Song et al.,
1994; Dahl et al., 2000). The sequence comparison revealed that the proteins have not much in common except the catalysis of pyroglutamyl formation. These findings prompted us to compare the substrate specificities
of a plant and human QC in detail, in order to identify differences of the catalysis. Papaya QC, representing the
first plant enzyme characterized, and human QC, representative for the highly homologous mammalian QCs,
were chosen for this comparison. Papaya QC was purified from crude papain and the human enzyme was recombinantly expressed in P. pastoris. The recombinant
QC was shown to have very similar characteristics and
an identical specific activity compared to the highly homologous QC purified from bovine pituitary, suggesting
that the recombinant enzyme resembles the native QC
very closely (Schilling et al., 2002a).
Due to difficulties in detecting pyroglutamyl formation
directly, coupled enzymatic assays have been developed. Detection of ammonia liberated in a coupled assay
with glutamate dehydrogenase is most suitable, since
ammonia is formed by conversion of nearly all QC substrates (Bateman, 1989). Originally, the assay was developed as a time-consuming discontinuous method. An increase in the amount of the auxiliary enzyme, however,
allowed a continuous data monitoring, suitable to determine the kinetic parameters for various substrates. With
the continuous assay, the kinetic parameters for conversion of about 40 glutaminyl peptides by plant and human
QC were determined. Both glutaminyl cyclases share obvious similarities in their catalytic properties (Table 1);
they exhibited the highest turnover numbers with substrates containing an aromatic amino acid residue in the
penultimate position, and similar specificities were observed for peptides that consist of more than two amino
acids. Furthermore, both QC forms revealed substrate inhibition only for the fluorogenic substrates, and showed
an overall similar dependence on the ionic strength. Finally, the pH dependence of the specificity constant
kcat/Km reveals a dependence of the overall catalysis on
an unprotonated substrate amino group, and both QC
forms catalyze the cyclization of L-βhomoglutaminyl
residues with identical competence. Obviously, the QCs
from plants and animals catalyze the cyclization of the
glutaminyl residue with very similar efficiency and probably by an identical overall mechanism, i.e. non-covalent
catalysis, by facilitating the intramolecular cyclization of
the glutaminyl residue. This was already suggested for
papaya QC (Gololobov et al., 1994). Initial data for the animal QC pointed to a nucleophilic influence by free thiol
group(s), and subsequent formation of an acyl-enzyme
during catalysis was suggested (Busby et al., 1987). More
recently, however, human QC was shown to carry no free
thiol groups, and the inhibition by iodoacetamide previously observed could be explained by a side reaction
(Bateman et al., 2001; Schilling et al., 2002a). Non-covalent catalysis by the human QC is also corroborated by
the fact that this protein is assumed to possess a fold
very similar to bacterial Zn-dependent aminopeptidases
and that QC might be evolved from an ancestral
aminopeptidase (Bateman et al., 2001). Interestingly, also
the related aminopeptidase from Aeromonas proteolytica
(AAP), like human QC, showed a remarkably improved
substrate specificity towards longer peptide substrates
1590
S. Schilling et al.
possibly due to interactions between the P1’- and P2’amino acid residue in the substrate and the enzyme
(Wilkes et al., 1973). Therefore, it is tempting to suggest
that animal QCs evolved from a Zn-dependent aminopeptidase keeping the secondary substrate interaction
sites, but underwent structural rearrangements of the active-site geometry accompanied by a switch in the catalytic mechanism.
In contrast to human QC, papaya QC did not show
such a distinct secondary specificity (Table 2). This holds
true for a relaxed selectivity pattern concerning substrate length and composition, as well as the different
kcat/Km-ratio for the partially cyclized substrate H-Gln-cyclo(Nε-Lys-Arg-Pro-Ala-Gly-Phe). Moreover, substrates
containing proline in the second position and the ester
substrate H-Gln-OtBu are well accepted by plant QC but
only weakly by human QC. A further striking difference is
the inability of human QC to convert the methylated glutaminyl residue of H-Gln(NMe)-Phe-Lys-Ala-Glu-NH2
into a pyroglutamyl residue. This property seems to be
attributable to N-terminal substrate binding, since this
peptide did not inhibit human QC. Furthermore, neither
cyclization of, nor inhibition by, the peptide H-Glu(NHNH2)-Ser-Pro-Thr-Ala-NH2 was detected for papaya QC.
Human QC, however, was inhibited competitively by the
N-terminal hydrazide peptide. The presented data suggest that both QCs differ in substrate binding. Obviously, there is more flexibility for substrate side-chain recognition by both enzymes. Interestingly, this is in great
contrast to the strict requirement of an unprotonated
substrate amino group for catalysis of both enzymes
(Figure 3).
The conversion of the potential physiological substrates of human QC [Gln1]-gastrin, [Gln1]-neurotensin
and [Gln1]-FPP, shown here for the first time, reflects the
relatively broad substrate specificity of human QC. Possibly, the QC is physiologically active in the pGlu formation of most if not all N-terminally Gln-containing hormones. This implies that QC may occur in more tissues
than detected to date (Pohl et al., 1991; Sykes et al.,
1999). The QC could play a key role in the formation of the
bioactive structure of pGlu-containing hormones.
Conclusion
The substrate specificity of human and papaya glutaminyl cyclase, enzymes that catalyze the same reaction
but are not related in sequence, was analyzed in a comparative manner. Both enzymes reveal some analogy in
their catalytic behavior, i.e., they show
(i) a dependence of catalysis on an unprotonated N-terminal substrate amino group,
(ii) a similar specificity towards oligopeptide substrates
and
(iii) similar requirements for the distance between αamino- and γ-amide group of the N-terminal glutaminyl residue.
Differences, however, are evident in
(iv) the conversion of short peptide substrates and
dipeptide surrogates,
(v) the dependence of size and amino acid composition
of the substrate and
(vi) the recognition mode of the modified γ-amide group
of the N-terminal glutamine of the substrate.
To our knowledge, this is first detailed analysis of substrate specificity of an animal QC and the first direct
comparison of a plant and an animal QC. The continuous
assay introduced allows the testing of a number of glutaminyl peptides revealing similarities in the overall potency for catalysis of glutamine cyclization, but also differences in the specificity of the QCs. Obviously, the
differences are mainly based on dissimilar secondary
binding sites for the substrates and a different recognition pattern of the N-terminal glutamine residue.
Materials and Methods
Materials
Papaya QC was purified from crude papain as described previously (Schilling et al., 2002b). Human QC was expressed in P.
pastoris and purified from the fermentation broth as described
(Schilling et al., 2002a). The QC substrates H-Gln-Gln-OH,
H-Gln-Glu-OH, H-Gln-Gly-OH, H-Gln-βNA, H-Gln-AMC and
H-Gln-Ala-OH were purchased from Bachem (Bubendorf,
Switzerland) or Senn Chemicals (Dielsdorf, Switzerland). Pyroglutamate aminopeptidase from Bacillus amyloliquefaciens was
supplied by Qiagen (Hilden, Germany). Glutamate dehydrogenase from bovine liver was purchased from Fluka (Seelze, Germany). NADH/H+ and α-ketoglutaric acid were obtained from
Sigma (St. Louis, USA). All other chemicals were of analytical or
HPLC grade.
Peptide Synthesis
Oligopeptides Peptides were synthesized semiautomatically
at a 0.5 mmol scale using a peptide synthesizer (Labortec
SP650, Bachem) as previously described (Schilling et al.,
2002a). Longer peptides were synthesized at a 25 µmol scale
using the automated Symphony peptide synthesizer (Rainin Instrument Co., Emeryville, USA) as described (Manhart et al.,
2003). For all peptide couplings, modified Fmoc-protocols of
solid-phase peptide synthesis were employed using 2-(1H-Benzotriazole-1-yl)-1,1,3,3,-tetramethyluronium tetrafluoroborate
(TBTU; Novabiochem, Schwalbach, Germany)/base (diisopropyl
ethylamine or N-methyl-morpholine; Merck, Darmstadt, Germany) or in case of difficult couplings N-[(dimethylamino)-1H1,2,3,-triazolo[4,5-b]pyridin-1-ylmethylene]-N-methylmethanammonium hexafluorophosphate N-oxide (4,5) (HATU;
Applied Biosystems, Foster City, USA)/diisopropyl ethylamine
as activating reagents were used. After cleavage from the resin
by a trifluoroacetic acid (TFA; Merck) containing cocktail, the
crude peptides were purified by preparative HPLC with acid free
solvents in order to avoid further cyclization of the N-terminal
glutamine. Preparative HPLC was performed with a linear gradient of acetonitrile in water (5 – 40% or 65% acetonitrile over
40 min) on a 250 – 21 Luna RP18 column (Phenomenex, Torrance, USA). To confirm peptide purity and identity analytical
HPLC and ESI-MS was applied.
Substrate Specificity of Glutaminyl Cyclase
H-Glu(NH-NH2)-Ser-Pro-Thr-Ala-NH2 The linear precursor
peptide (Fmoc-Glu-Ser-Pro-Thr-Ala-NH2) was synthesized according to a standard Fmoc-procedure (Schilling et al., 2002a)
on Rink amide MBHA resin (Novabiochem). After cleavage of the
Fmoc-protected peptide from the resin, the peptide was precipitated with diethyl ether, filtered and dried. HMBA-AM resin
(1.16 mmol/g, Novabiochem) was used for coupling of the γ-carboxylic acid group of glutamic acid of the precursor peptide
(3 eq.) in dichloromethane (DCM, Merck). Dicyclohexylcarbodiimide (DCC, Serva, Heidelberg, Germany) (4 eq.) and dimethylaminopyridine (DMAP, Aldrich, St. Louis, USA) (0.1 eq.) were
used as coupling reagents. After 12 hours the resin was filtered,
washed with DCM and the reaction was repeated. After deprotection of the N-terminal Fmoc-group by employing 20% piperidine in dimethylformamide (DMF) (3×5 min) the peptide resin
was treated with a 5% hydrazine solution (20 ml/g) for 1.5 hours.
The resin was filtered and washed with DMF and TFA. Following
evaporation, the crude peptide was precipitated with ether giving 76% yield.
H-Gln-Lys(Gln)-Arg-Leu-Ala-NH2 The linear peptide was
synthesized according to standard Fmoc-procedure on Rink
amide MBHA resin (Schilling et al., 2002a) using FmocLys(Fmoc)-OH as penultimate amino acid coupling. After deprotection of the two amino protecting groups of lysine with 20%
piperidine (Merck) in DMF, 4 eq. Fmoc-Gln(Trt)-OH were coupled. Standard cleavage procedure resulted in 95% yield.
H-Gln(NMe)-Phe-Lys-Ala-Glu-NH2 Fmoc-Gln(NMe)-OH was
synthesized starting from Fmoc-Glu-OtBu loaded on Fmoc-MIAM (Novabiochem) resin. After swelling with DCM, the resin (0.5
g) was washed with DMF and deprotected with 20% piperidine
solution in DMF. The resin was given into 5 ml DMF and 5 eq.
Fmoc-Glu-OtBu, 5 eq. HATU and 10 eq. diisopropyl ethylamine
were added subsequently and shaken for 6 hours. After filtration
and washing, the product was cleaved according to standard
TFA cleavage conditions. The peptide H-Gln(NMe)-Phe-LysAla-Glu-NH2 was synthesized as described (Schilling et al.,
2002a). Fmoc-Gln(NMe)-OH was coupled with HATU/diisopropyl ethylamine overnight. The standard cleavage procedure
resulted in a yield of 78% of the crude peptide.
H-Glu(OMe)--naphthylamide, H-Gln-Val-OH, H-Gln-Tyr-OH
Boc-protected dipeptides were synthezised applying standard
mixed anhydride procedure by using isobutyl chlorocarbonate
(Merck). The C-terminal methylesters Boc-Gln-Tyr-OMe and
Boc-Gln-Val-OMe were saponified by 1 N NaOH in dioxane. The
Boc-protected peptides were deprotected by HCl/dioxane solution for 10 min. After evaporation the residue was crystallized
with several solvents giving 60 – 70% of a solid compound.
H-Gln-cyclo(N-Lys-Arg-Pro-Ala-Gly-Phe) The linear precursor Boc-Gln(Trt)-Lys-Arg(Pmc)-Ala-Gly-Phe-OH was synthesized on acid sensitive 2-chlorotrityl resin. Coupling was carried
out using a standard protocol of Fmoc-strategy using FmocLys(Mtt)-OH. After cleavage with 3% TFA solution in DCM (10
times 5 min), the solution was neutralized with 10% pyridine in
methanol (MeOH; Merck), washed 3 times with DCM and MeOH,
evaporated to 5% of the volume and the crude peptide was precipitated with ice-cold water. Following, the crude peptide was
cyclized using DCC/N-hydroxybenzotriazole (HOBt; Aldrich) activation. The crude peptide was dissolved in dry dichloromethane (0.2 mmol/50 ml), 0.2 mmol N-methylmorpholine and
0.4 mmol 1-hydroxybenzotriazole were added. This solution was
added dropwise to a solution of 0.4 mmol dicyclohexylcarbodiimide in 250 ml dichloromethane at 0 °C. The reaction was com-
1591
pleted by stirring overnight at room temperature. After filtration
of N,N´-dicyclohexylurea, the solvent was removed by evaporation. The residue was dissolved in ethyl acetate and washed several times with 1 N HCl, saturated solution of NaHCO3 and water.
The solution was dried over anhydrous Na2SO4, filtered and
evaporated to dryness in vacuum. The crude product was purified by RP-HPLC yielding 12% of the cyclic peptide.
Fluorometric Assays of QC
Human QC activity was evaluated using H-Gln-βNA at 30 °C, essentially as described (Schilling et al., 2002b). Briefly, the samples consisted of 0.2 mM fluorogenic substrate, 0.1 U pyroglutamyl aminopeptidase in 0.05 M Tris/HCl, pH 8.0 containing 5 mM
EDTA and an appropriately diluted aliquot of QC in a final volume
of 250 µl. In case of papaya QC, the Tris buffer was substituted
by 0.05 M Tricine/NaOH, pH 8.0. Excitation/emission wavelength
was 320/405 nm. The assay reactions were initiated by addition
of QC. Enzymatic activity was determined from the amount of released βNA calculated using a standard curve of βNA under assay conditions. One Unit is defined as the amount of QC
catalysing the formation of 1 µmol pGlu-βNA from H-Gln-βNA
per minute under the described conditions. Reaction conditions
were the same in case of H-Gln-AMC, except that the excitation/emission wavelength was adjusted to 380/460 nm. The
measurements were performed with a Novostar (BMG Labtechnologies, Offenburg, Germany) or a SpectraFluor Plus reader for
microplates (Tecan, Maennedorf, Switzerland).
Spectrophotometric Assay of QC
QC activity was analyzed spectrophotometrically using a continuous assay that was established by adapting a discontinuous
method (Bateman, 1989) using glutamate dehydrogenase as
auxiliary enzyme. Samples consisted of the respective QC substrate, 0.3 mM NADH, 14 mM α-ketoglutaric acid and 30 U/ml
glutamate dehydrogenase in a final volume of 250 µl. Reactions
were started by addition of QC and pursued by monitoring of the
decrease in absorbance at 340 nm for 8 – 15 min. The initial velocities were evaluated and the enzymatic activity was determined from a standard curve of ammonia obtained previously
under assay conditions. All samples were measured at 30 °C, using either the SpectraFluor Plus or the Sunrise reader for microplates (both from Tecan). Kinetic data were evaluated using
GraFit software (version 5.0.4. for windows, Erithacus Software
Ltd., Horley, UK).
pH Dependence
The specificity rate constants (kcat/Km) at varying pH values were
determined under first-order conditions, i.e., at substrate concentrations lower than Km, using H-Gln-βNA as substrate. The
reactions were measured either in a three-component buffer
system that provides a constant ionic strength over a wide pHrange consisting of 0.025 M MES, 0.025 M acetic acid and 0.05 M
Tris, or in 0.05 mM MOPS (pH 6.2 – 8.2) buffer at a constant ionic
strength of 0.05 M, adjusted by addition of NaCl. The buffers
were titrated to the desired pH using HCl or NaOH. All measurements were performed with the Novostar reader (BMG Labtechnologies) for microplates at 23 °C.
The pH-dependent rate data were evaluated by nonlinear regression utilising GraFit software. Measured rates were fitted to
the following equation:
kcat/Km = kcat/Km(limit)[1/(1 + 10pKI-pH + 10pH-pKII)],
1592
S. Schilling et al.
in which KI and KII are the dissociation constants of the catalytically important functional groups and kcat/Km(limit) is the pH-independent maximum rate constant.
MALDI-TOF Mass Spectrometry
Matrix-assisted laser desorption/ionization mass spectrometry
was carried out using the Hewlett-Packard G2025 LD-TOF System (Palo Alto, USA) with a linear time-of-flight analyzer. The instrument was equipped with a 337 nm nitrogen laser, a potential
acceleration source (5 kV) and a 1.0 m flight tube. Detector operation was in the positive-ion mode and signals were recorded
and filtered using LeCroy 9350M digital storage oscilloscope
linked to a personal computer. Samples (5 µl) were mixed with
equal volumes of the matrix solution. For matrix solution we
used DHAP/DAHC, prepared by solving 30 mg 2´,6´-dihydroxyacetophenone (Aldrich) and 44 mg diammonium hydrogen citrate (Fluka) in 1 ml acetonitrile/0.1% TFA in water (1/1, v/v). A
small volume (≈1 µl) of the matrix-analyte-mixture was transferred to a probe tip and immediately evaporated in a vacuum
chamber (Hewlett-Packard G2024A sample prep accessory) to
ensure rapid and homogeneous sample crystallization.
The enzymatic reactions were performed in samples of 100 µl
consisting of QC (0.01 – 1 U) and 0.5 mM substrate in 0.04 M
Tris/HCl, pH 7.7, at 30 °C. At the times indicated, aliquots were
removed, diluted with matrix and analyzed as described.
Acknowledgments
We thank Dr. J.-U. Rahfeld for QC modeling and Dr. F. Rosche for
the help in recording the mass spectra. The technical assistance
of H. Cynis and J. Zwanzig is gratefully acknowledged. We thank
Dr. S.A. Hinke for critical reading of the manuscript.
References
Bateman, R. C. J. (1989). A spectrophotometric assay for glutaminyl-peptide cyclizing enzymes. J. Neurosci. Methods 30,
23 – 28.
Bateman, R. C., Temple, J. S., Misquitta, S. A., and Booth, R. E.
(2001). Evidence for essential histidines in human pituitary
glutaminyl cyclase. Biochemistry 40, 11246 – 11250.
Busby, W. H. J., Quackenbush, G. E., Humm, J., Youngblood, W.
W., and Kizer, J. S. (1987). An enzyme(s) that converts glutaminyl-peptides into pyroglutamyl-peptides. Presence in pituitary, brain, adrenal medulla, and lymphocytes. J. Biol. Chem.
262, 8532 – 8536.
Consalvo, A. P., Young, S. D., Jones, B. N., and Tamburini, P. P.
(1988). A rapid fluorometric assay for N-terminal glutaminyl
cyclase activity using high-performance liquid chromatography. Anal. Biochem. 175, 131 – 138.
Dahl, S. W., Slaughter, C., Lauritzen, C., Bateman, R. C. Jr., Connerton, I., and Pedersen, J. (2000). Carica papaya glutamine
cyclotransferase belongs to a novel plant enzyme subfamily:
cloning and characterization of the recombinant enzyme. Protein Expr. Purif. 20, 27 – 36.
El Moussaoui, A., Nijs, M., Paul, C., Wintjens, R., Vincentelli, J.,
Azarkan, M., and Looze, Y. (2001). Revisiting the enzymes
stored in the laticifers of Carica papaya in the context of their
possible participation in the plant defence mechanism. Cell
Mol. Life Sci. 58, 556 – 570.
Ellis, K. J. and Morrison, J. F. (1982). Buffers of constant ionic
strength for studying pH-dependent processes. Methods Enzymol. 87, 405 – 426.
Fischer, W. H. and Spiess, J. (1987). Identification of a mammalian glutaminyl cyclase converting glutaminyl into pyroglutamyl peptides. Proc. Natl. Acad. Sci. USA 84, 3628 – 3632.
Gololobov, M. Y., Song, I., Wang, W., and Bateman, R. C. J.
(1994). Steady-state kinetics of glutamine cyclotransferase.
Arch. Biochem. Biophys. 309, 300 – 307.
Gololobov, M. Y., Wang, W., and Bateman, R. C. J. (1996). Substrate and inhibitor specificity of glutamine cyclotransferase
(QC). Biol. Chem. Hoppe-Seyler 377, 395 – 398.
Manhart, S., Hinke, S. A., McIntosh, C. H., Pedersen, J., and Demuth, H.-U. (2003). Structure-function analysis of a series of
novel GIP analogues containing different helical length linkers.
Biochemistry 42, 3081 – 3088.
McClure, W. R. (1969). A kinetic analysis of coupled enzyme assays. Biochemistry 8, 2782 – 2786.
Messer, M. (1963). Enzymatic cyclization of L-glutamine and
L-glutaminyl peptides. Nature 4874, 1299
Oberg, K. A., Ruysschaert, J. M., Azarkan, M., Smolders, N.,
Zerhouni, S., Wintjens, R., Amrani, A., and Looze, Y. (1998).
Papaya glutamine cyclase, a plant enzyme highly resistant to
proteolysis, adopts an all-β conformation. Eur. J. Biochem.
258, 214 – 222.
Pohl, T., Zimmer, M., Mugele, K., and Spiess, J. (1991). Primary
structure and functional expression of a glutaminyl cyclase.
Proc. Natl. Acad. Sci. USA 88, 10059 – 10063.
Schilling, S., Hoffmann, T., Rosche, F., Manhart, S., Wasternack,
C., and Demuth, H. U. (2002a). Heterologous expression and
characterization of human glutaminyl cyclase: evidence for a
disulfide bond with importance for catalytic activity. Biochemistry 41, 10849 – 10857.
Schilling, S., Hoffmann, T., Wermann, M., Heiser, U., Wasternack, C., and Demuth, H.-U. (2002b). Continuous spectrometric assays for glutaminyl cyclase activity. Anal. Biochem.
303, 49 – 56.
Song, I., Chuang, C. Z., and Bateman, R. C. J. (1994). Molecular
cloning, sequence analysis and expression of human pituitary
glutaminyl cyclase. J. Mol. Endocrinol. 13, 77 – 86.
Sykes, P. A., Watson, S. J., Temple, J. S., and Bateman, R. C. J.
(1999). Evidence for tissue-specific forms of glutaminyl cyclase. FEBS Lett. 455, 159 – 161.
Tsuru, D., Fujiwara, K., and Kado, K. (1978). Purification and
characterization of L-pyrrolidonecarboxylate peptidase from
Bacillus amyloiliquefaciens. J. Biochem. (Tokyo) 84, 467 – 476.
Wilkes, S. H., Bayliss, M. E., and Prescott, J. M. (1973). Specificity of aeromonas aminopeptidase toward oligopeptides
and polypeptides. Eur. J. Biochem. 34, 459 – 466.
Zerhouni, S., Amrani, A., Nijs, M., Smolders, N., Azarkan, M.,
Vincentelli, J., and Looze, Y. (1998). Purification and characterization of papaya glutamine cyclotransferase, a plant enzyme highly resistant to chemical, acid and thermal denaturation. Biochim. Biophys. Acta 1387, 275 – 290.
Received July 29, 2003; accepted October 7, 2003
THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2003 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 278, No. 50, Issue of December 12, pp. 49773–49779, 2003
Printed in U.S.A.
Identification of Human Glutaminyl Cyclase as a Metalloenzyme
POTENT INHIBITION BY IMIDAZOLE DERIVATIVES AND HETEROCYCLIC CHELATORS*
Received for publication, August 15, 2003, and in revised form, September 26, 2003
Published, JBC Papers in Press, September 30, 2003, DOI 10.1074/jbc.M309077200
Stephan Schilling‡, André J. Niestroj‡, Jens-Ulrich Rahfeld‡, Torsten Hoffmann‡,
Michael Wermann‡, Katrin Zunkel‡, Claus Wasternack§, and Hans-Ulrich Demuth‡¶
From the ‡Probiodrug Aktiengesellschaft, Weinbergweg 22, and the §Leibniz Institute for Plant Biochemistry,
Post Office Box 110432, 06120 Halle/Saale, Germany
Human glutaminyl cyclase (QC) was identified as a
metalloenzyme as suggested by the time-dependent inhibition by the heterocyclic chelators 1,10-phenanthroline and dipicolinic acid. The effect of EDTA on QC
catalysis was negligible. Inactivated enzyme could be
fully restored by the addition of Zn2ⴙ in the presence of
equimolar concentrations of EDTA. Little reactivation
was observed with Co2ⴙ and Mn2ⴙ. Other metal ions
such as Kⴙ, Ca2ⴙ, and Ni2ⴙ were inactive under the same
conditions. Additionally, imidazole and imidazole derivatives were identified as competitive inhibitors of QC.
An initial structure activity-based inhibitor screening of
imidazole-derived compounds revealed potent inhibition of QC by imidazole N-1 derivatives. Subsequent
data base screening led to the identification of two
highly potent inhibitors, 3-[3-(1H-imidazol-1-yl)propyl]2-thioxoimidazolidin-4-one and 1,4-bis-(imidazol-1-yl)methyl-2,5-dimethylbenzene, which exhibited respective Ki values of 818 ⴞ 1 and 295 ⴞ 5 nM. The binding
properties of the imidazole derivatives were further analyzed by the pH dependence of QC inhibition. The kinetically obtained pKa values of 6.94 ⴞ 0.02, 6.93 ⴞ 0.03,
and 5.60 ⴞ 0.05 for imidazole, methylimidazole, and
benzimidazole, respectively, match the values obtained
by titrimetric pKa determination, indicating the requirement for an unprotonated nitrogen for binding to
QC. Similarly, the pH dependence of the kinetic parameter Km for the QC-catalyzed conversion of H-Gln-7-amino-4-methylcoumarin also implies that only N-terminally unprotonated substrate molecules are bound to
the active site of the enzyme, whereas turnover is not
affected. The results reveal human QC as a metal-dependent transferase, suggesting that the active sitebound metal is a potential site for interaction with
novel, highly potent competitive inhibitors.
Glutaminyl cyclases (QC)1 (EC 2.3.2.5) are acyltransferases
present in animals and plants that catalyze the conversion of
N-terminal glutaminyl residues into pyroglutamic acid with
the concomitant liberation of ammonia (Scheme 1). Several
peptide hormones and proteins carry N-terminal pyroglutamyl
residues. Previously, the formation of N-terminal pyrogluta-
* The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
¶ To whom correspondence should be addressed. Tel.: 49-3455559900; Fax: 49-345-5559901; E-mail: hans-ulrich.demuth@probiodrug.
de.
1
The abbreviations used are: QC, glutaminyl cyclase; AMC, 7-amino4-methylcoumarin; Mes, 4-morpholineethanesulfonic acid.
This paper is available on line at http://www.jbc.org
mate from glutamine was assumed to proceed spontaneously
(1). However, the QCs were identified more recently as catalysts of the reaction in both mammals and plants (2–5). Generally, QCs from both mammalian and plant sources appear to
be very similar monomeric proteins that are expressed in the
secretory pathways and have similar molecular masses, ⬃33
and ⬃40 kDa, respectively (6, 7). Their primary structures,
however, reveal no sequence homology, and the enzymes feature completely different folding patterns. Whereas plant QC
consists almost solely of ␤-sheet structure, mammalian QCs
are predicted to possess an ␣/␤-fold (8 –10). Furthermore, plant
QC does not share sequence or structural homology to other
plant enzymes, belonging, apparently, to a separate enzyme
subfamily (4). Mammalian QCs, however, exhibit remarkable
homology toward bacterial aminopeptidases, suggesting their
evolutionary origin in this protein family (9).
Investigating the substrate specificity of both enzymes, we
recently found differences between papaya and human QC (24).
Although the catalysis of cyclization of and the inhibition by
modified N-terminal residues were quite different, both enzymes catalyzed the cyclization of oligopeptides with similar
specificity rate constants. Moreover, they also catalyzed the
formation of a five-membered lactam ring from L-␤-homoglutaminyl peptides. Based on the present state of knowledge, it is
assumed that plant and animal QCs catalyze the formation of
N-terminal pyroglutamic acid (pGlu) residues by enabling the
intramolecular cyclization of the glutaminyl residue in a noncovalent manner (11, 12). However, the results of the substrate
specificity study revealed differences in substrate recognition
by both enzymes.
Thus, a more detailed analysis of the inhibition pattern of
plant and human QCs should help to deepen our understanding of QC catalysis. To date, however, systematic inhibitor data
exist only for plant QC, which is competitively inhibited by
peptides containing an N-terminal proline residue (13),
whereas human QC was shown to be inactivated by 1,10phenanthroline and reduced 6-methylpterin (3). Thus, we investigated the inhibiting potency of heterocyclic compounds
from different structural classes. Among them, imidazole and
structurally related compounds were found to be the most
efficient competitive inhibitors of human QC. The data provide
the first insights into enzyme/inhibitor interactions, offer clues
for further optimization of the inhibitory structure, and reveal
novel aspects of human QC catalysis.
EXPERIMENTAL PROCEDURES
Materials—Human QC was recombinantly expressed in Pichia
pastoris and purified as described previously (10). Chemicals were
purchased as follows. Glutamate dehydrogenase was from Fluka, pyroglutamyl aminopeptidase came from Qiagen, H-Gln-AMC and H-GlnGln-OH were from Bachem, NADH/H⫹ and ␣-ketoglutaric acid were
from Sigma, and the imidazole derivatives were from Acros, Sigma-
49773
49774
Competitive Inhibition of Metal Enzyme Glutaminyl Cyclase
TABLE I
Inhibition constants of imidazole derivatives in the
human QC catalyzed reaction
Assays were performed at 30 °C in 0.05 M Tris-HCl, pH 8.
Compound
Ki value
␮M
SCHEME 1. N-terminal cyclization of glutaminyl peptides by QC.
Core structures
Imidazole
Benzimidazole
N-1 derivatives
1-Benzylimidazole
1-Methylimidazole
1-Vinylimidazole
Oxalic acid diimidazolidide
N-Acetylimidazole
N-(Trimethylsilyl)-imidazole
N-Benzoylimidazole
1-(2-Oxo-2-phenylethyl)-imidazole
1-(3-Aminopropyl)-imidazole
1-Phenylimidazole
1,1⬘-Sulfonyldiimidazole
C-4 (5) derivatives
N-␻-acetylhistamine
L-Histidinamide
H-His-Trp-OH
L-Histidinol
L-Histidine
4-Imidazole-carboxaldehyde
Imidazol-4-carbonic acid methylester
C-4,5 derivatives
5-Hydroxymethyl-4-methyl-imidazole
5-Amino-3H-imidazole-4-carboxylic acid amide
4,5-Diphenyl-imidazole
4,5-Dicyanoimidazole
C-2 derivatives
2-Methyl-benzylimidazole
2-Ethyl-4-methyl-imidazole
2-Aminobenzimidazole
2-Chloro-1H-benzimidazole
103 ⫾ 4
138 ⫾ 5
7.1 ⫾ 0.3
30 ⫾ 1
49 ⫾ 2
78 ⫾ 2
107 ⫾ 3
167 ⫾ 7
174 ⫾ 7
184 ⫾ 5
410 ⫾ 10
No inhibition
No inhibition
17 ⫾ 1
560 ⫾ 40
600 ⫾ 30
1530 ⫾ 120
4400 ⫾ 200
7600 ⫾ 700
14500 ⫾ 600
129 ⫾ 5
15500 ⫾ 500
No inhibition
No inhibition
165 ⫾ 4
580 ⫾ 40
1800 ⫾ 100
No inhibition
FIG. 1. Lineweaver-Burk plots for human QC catalyzed cyclization of H-Gln-AMC in presence of various concentrations of
imidazole between 0.03 and 1 mM. The inset shows a secondary plot
of the obtained slopes of the Lineweaver-Burk evaluation versus the
inhibitor concentrations. The conditions were 0.05 M Tris/HCl, pH 8.0,
containing 5 mM EDTA at 30 °C.
Aldrich, Asinex, TimTec, and InterBioScreen. Papaya QC was purified
from crude papain as described previously (14). All other chemicals
were of analytical or high pressure liquid chromatography grade.
QC Assays—QC activity was assayed essentially as described elsewhere (14). Briefly, the assay reactions (250 ␮l) consisted of varying
concentrations (⬃0.25– 4 Km) of H-Gln-AMC or H-Gln-Gln-OH in 0.05 M
Tris-HCl, pH 8.0. In some cases, assay buffer also contained up to 5 mM
EDTA, which has been shown to stabilize the auxiliary enzyme (15). Ki
values of QC inhibition were not influenced by EDTA. In the case of the
spectrophotometric assay, samples additionally contained 30 units/ml
glutamic acid dehydrogenase, 0.25 mM NADH/H⫹, and 15 mM ␣-ketoglutaric acid. Reactions were started by the addition of QC, and activity
was monitored by recording the decrease in absorbance at 340 nm. In
contrast, the assay samples for the fluorometric detection of QC activity
contained only buffer and 0.4 units/ml pyroglutamyl aminopeptidase as
an auxiliary enzyme. The excitation/emission wavelengths were 380/
460 nm. Reactions were started by the addition of QC. QC activity was
determined from a standard curve of AMC under assay conditions. All
determinations were carried out at 30 °C using either the TECAN
SpectraFluor Plus or the BMG Novostar reader for microplates.
Inhibitor Assay—For inhibitor testing, the sample composition was
the same as described above, except for the addition of the inhibitory
compound. For rapid inhibitor screening, samples contained up to 4 mM
of the respective imidazole derivative and a substrate concentration
equal to the Km value of the test substrate. For detailed investigations
of the inhibition and determination of Ki values, the influence of the
inhibitor on the auxiliary enzymes was investigated first. In no case
was an influence on one of the auxiliary enzymes detected, enabling the
reliable determination of the QC inhibition. The inhibition constants
were evaluated by fitting the data of the obtained progress curves
according to the general equation for competitive inhibition using
SCHEME 2. The constitution of the imidazole ring (A) and an imidazole
N-1 derivative (B).
GraFit software (version 5.0.4. for windows, Erithacus Software Ltd.,
Horley, UK).
pH Dependence—For the investigation of the pH dependence of QC
catalysis and inhibition, the previously developed fluorometric assay
was used (14). Determinations were carried out at 30 °C in a buffer
originally used by Ellis and Morrison consisting of 0.06 M acetic acid,
0.06 M Mes, and 0.12 M Tris (16). The buffer provides a constant ionic
strength over the entire pH range chosen. Additionally, the activity of
human QC acting on H-Gln-AMC has shown to be quite independent
from variations in ionic strength. The resulting pH dependence data
were fitted to single dissociation models for the inhibitors or to equations that account for two dissociating groups in the case of the kinetic
parameter Km using Grafit software. Because of the reduced stability
and catalytic activity of the auxiliary enzyme pyroglutamyl aminopeptidase under acidic and basic conditions, the study was limited to the
range between 5.5 and 8.5 pH.
Enzyme Inactivation/Reactivation Procedures—An aliquot of human
QC (0.1– 0.5 mg, 1 mg/ml) was inactivated overnight by dialysis against
a 3000-fold excess of 5 mM 1,10-phenanthroline, or 5 mM dipicolinic acid
in 0.05 M Bis-Tris/HCl, pH 6.8. Subsequently, the inactivating agent
was carefully removed by dialysis (three cycles, 2000-fold excess) of the
samples against 0.05 M Bis-Tris/HCl, pH 6.8, containing 1 mM EDTA.
Competitive Inhibition of Metal Enzyme Glutaminyl Cyclase
49775
TABLE II
Inhibition of human QC by N1 derivatives selected from compound databases
Determinations were carried out as described in Table I.
Reactivation experiments were performed at room temperature for 15
min using Zn2⫹, Mn2⫹, Ni2⫹, Ca2⫹, K⫹, and Co2⫹ ions at concentrations
of 1, 0.5, and 0.25 mM in 0.025 M Bis-Tris, pH 6.8, containing 0.5 mM
EDTA. QC activity assays were performed in 0.05 M Tris/HCl, pH 8,
containing 2 mM EDTA to avoid a rapid reactivation by the traces of
metal ions present in buffer solutions.
RESULTS
Inhibition by Imidazole—Because neither glutamic acid dehydrogenase nor pyroglutamyl aminopeptidase were inhibited
by imidazole in the concentration range used, both the fluorometric as well as the spectrophotometric assay could be applied. The
Lineweaver-Burk plot of the fluorometric assay data (Fig. 1)
reveals competitive inhibition by imidazole (inset in Fig. 1). Thus,
imidazole binds in the active site completely blocking substrate
conversion. The Ki values of 103 ⫾ 4 and 129 ⫾ 8 ␮M obtained
with the fluorometric and chromogenic assay, respectively,
match very well. Interestingly, benzimidazole inhibits human
QC similarly as does imidazole, also exhibiting linear competition and a Ki value of 138 ⫾ 5 ␮M. Obviously, the condensed ring
structure does not influence significantly the binding of the compound under the conditions used. Driven by the similar inhibitory characteristics of imidazole and benzimidazole, derivatives
of both compounds were tested to identify secondary interactions
that improve or diminish their inhibitory potency.
Imidazole Derivatives—Imidazole and benzimidazole derivatives carrying substituents at different positions of the ring
system were tested, and the influence of the substituents was
compared (Table I). The constitution refers to the imidazole
ring (Scheme 2).
C-4 (5) and C-4,5 Derivatives—The compounds with substitutions either in the constitutionally equivalent 4 or 5 position
of the imidazole ring or in both positions showed reduced inhibitory activity toward human QC. In contrast, N-␻-acetylated histamine proved to be a potent inhibitory compound.
Small substituents in both positions (4 and 5) seemed to have
only minor effects on binding, as indicated by the similar inhibition constants of 5-hydroxymethyl-4-methyl-imidazole and imidazole itself. Larger and more bulky groups attached to these sites
diminished or abolished binding of the compounds to the enzyme.
However, some of the substituents of the tested imidazole derivatives are known to exert negative inductive or mesomeric effects, thereby reducing the electron density within the imidazole
ring. This could also contribute to poorer binding. The different
Ki values detected for L-histidine and histidinamide also indicate
an influence of the charge of the inhibitors on binding. Evidence
for electrostatic repulsion of charged substrates was observed
previously during an investigation of the substrate specificity of
QC, i.e. glutaminamide was readily converted to products by
human QC, but glutamine was not cyclized (24).
C-2 derivatives—All derivatives tested showed a diminished
binding to the active site of QC relative to imidazole. Obviously,
there is a strong impact on proper binding by any additional
atom in this position. For instance, the simple addition of a
methyl group to form 2-methyl-benzylimidazole reduces the
49776
Competitive Inhibition of Metal Enzyme Glutaminyl Cyclase
TABLE III
Inhibition of human QC by L-histamine and two derivatives
Assays were carried out as described in Table I. Another designation of the derivatives is tele-methylhistamine. They are in vivo occuring
metabolites of histamine.
inhibition constant of the interaction by about one order of
magnitude. A very similar relation becomes evident comparing
the Ki values for benzimidazole and 2-amino-benzimidazole.
N-1 Derivatives—Among the imidazole derivatives tested as
inhibitors of human QC, most compounds that had reduced Ki
values compared with imidazole contained modifications at the
N-1 nitrogen atom (Table I). Interestingly, only minor changes
in the N-substituent were necessary for substantial loss of
inhibitory power. This can be seen when comparing 1-benzylimidazole, 1-benzoylimidazole, and phenylimidazole as QC inhibitors. The data suggest, however, that steric hindrance for QC
binding of N-1 derivatives is marginal, opening up the possibility
for the development of even more potent QC inhibitors by structure optimization of N-1 modified imidazole compounds.
Compound Data Base Screening—The apparent improvement of the inhibitory power obtained by N-1 substitutions of
the imidazole ring allowed us to identify highly potent inhibitors of QC by data base screening. Some of the most potent
inhibitors are shown in Table II. In fact, the observed inhibition
constants are one order of magnitude lower as compared with
those determined in the initial structure-activity relationship
experiments (Table I). This approach led finally to the identification of hit compounds exhibiting Ki values of the QC inhibition in the nM range.
Effect of 1,4 and 1,5 Derivatization—The inhibition constants obtained for the 4(5)-substituted imidazole derivatives
already indicated that there are restrictions for efficient binding to the enzyme. An individual contribution of position 4 and
5, however, were undetectable, because both are identical with
respect to substitutions at one carbon. The individual effect of
substitutions in position 4 and 5 was analyzed by comparing
the inhibitory constants of L-histamine and the two intermediates in the biological degradation of histamine, 1-methyl-4histamine, and 1-methyl-5-histamine (Table III). Interestingly,
whereas methylation of one nitrogen of histamine forming
1-methyl-5-histamine improved the inhibitory activity considerably, methylation of the other nitrogen (1-methyl-4-histamine) led to a near total loss of inhibitory potential. Thus, steric
hindrance by the carbon atom adjacent to the basic nitrogen
seems to occur, which provides further support for the key role
of the basic nitrogen in binding of the imidazole derivatives to
the enzyme.
FIG. 2. The pH-dependence of the inhibition constant of imidazole (circles), benzimidazole (triangles), and 1-methylimidazole (squares) inhibiting human QC catalyzed cyclization of
H-Gln-AMC. Data points were fitted to a single dissociation model and
revealed pKa values of 6.94 ⫾ 0.02, 6.93 ⫾ 0.03 and 5.60 ⫾ 0.05 for
imidazole, 1-methylimidazole, and benzimidazole, respectively. Reactions were carried out at 30 °C in 0.06 M acetic acid, 0.06 M Mes, and
0.12 M Tris adjusted to the respective pH by the addition of NaOH
or HCl.
pH Dependence—The role of the basic nitrogen of imidazole
was further characterized through an investigation of the pHdependence of QC inhibition. Because of a limited catalytic
activity as well as the reduced stability of the auxiliary enzyme
pyroglutamyl aminopeptidase, this analysis was limited to a
pH range between pH 5.5 and 8.5. Because imidazole has a pKa
value near neutrality, however, this range was assumed to be
sufficiently wide for inspecting the influence of protonation and
deprotonation of the inhibitor. The inhibition of the QC-catalyzed reaction showed a strict dependence on the pH value (Fig.
2). With decreasing pH, the Ki-value of imidazole increased
drastically, exhibiting a 25-fold increase when moving from pH
8 to 5.5. Furthermore, in the basic pH region, Ki was constant,
suggesting that the potency of QC inhibition depends on deprotonation of the imidazole derivatives. This was also corroborated by fitting of the data to a single dissociation model (Fig.
2). The dissociating group influencing the inhibitory properties
of imidazole was characterized by a pKa value that is in excellent agreement with the pKa of the basic nitrogen of imidazole
(Table IV). Similar pH dependences were obtained for QC in-
Competitive Inhibition of Metal Enzyme Glutaminyl Cyclase
49777
TABLE IV
Comparison of the dissociation constants of inhibitors and a substrate determined by the pH-dependence of inhibitory and
Michelis-Menten constants as well as by acid/base titration
Assays were carried out at 30 °C in 0.06 M acetic acid, 0.06 M Mes, and 0.12 M Tris. The turnover number was found to be constant between pH
5.5 and 8.5. KM(I) reflects the dissociation of the substrate. KM(II) reveals a group of the enzyme.
Compound
Parameter
pKa kinetic determination
pKa titrimetric determination
Imidazole
Benzimidazole
1-Methylimidazole
H-Gln-AMC
Ki
Ki
Ki
KM(I)
KM(II)
6.94 ⫾ 0.02
5.60 ⫾ 0.05
6.93 ⫾ 0.03
6.81 ⫾ 0.04
8.60 ⫾ 0.10
6.946 ⫾ 0.003
5.500 ⫾ 0.010
7.000 ⫾ 0.003
6.830 ⫾ 0.010
—
FIG. 3. Inhibition of human QC by metal-chelating reagents.
Concentration dependence of inhibition by 1,10-phenanthroline (circles)
and EDTA (triangles) is shown. Residual activity of QC in the presence
of either compound was determined directly after the addition (dotted
traces) or preincubation of QC with the respective reagent for 15 min at
30 °C (continuous line).
hibition by benzimidazole and 1-methylimidazole (Fig. 2). For
both compounds, the kinetically determined pKa values compare well with the pKa values determined by titration (Table
IV). The dependence of the kinetic parameters Km and kcat on
the pH-value was also analyzed (data not shown). Whereas the
turnover number for conversion of H-Gln-AMC was not affected in the pH range between 5.5 and 8.5, the Michaelis
constant showed a simple pH dependence with an optimum
between pH 7.5– 8, tending to increase in the acidic and basic
pH region. The pH dependence of Km revealed a slope of 1 in the
acidic pH range, reflecting the presence of a single underlying
dissociative group. The kinetically determined pKa value in the
acidic range was nearly identical to the pKa of the substrate
H-Gln-AMC (Table IV). In the basic pH range, data revealed a
pKa value of 8.6 ⫾ 0.1, apparently reflecting a dissociating
group of the QC protein.
Inhibition by Heterocyclic Chelators—The inhibition of porcine QC by 1,10-phenanthroline has already been described (3).
However, the fact that EDTA has been shown to have an
activating effect on QC catalysis suggested that inhibition by
phenanthroline is not due to metal chelation (3). Also, in addition to being inhibited by 1,10-phenanthroline, human QCcatalyzed substrate cyclization was abolished in presence of
dipicolinic acid, another inhibitor of metalloenzymes. Both chelators inhibited QC in a competitive and time-dependent manner, i.e. initial activity that was already competitively inhibited
was found to be further reduced after prolonged incubation
with the compounds (Fig. 3). Interestingly, EDTA did not show
remarkable inhibition regardless of incubation time or under
any conditions.
Human QC was almost completely inactivated after extensive dialysis against 5 mM 1,10-phenanthroline or 5 mM dipicolinic acid. After repeated dialysis overnight against chelatorfree buffer solutions, QC activity was partially reactivated up
to 50 – 60% (data not shown). However, when dialyzed against
buffers containing 1 mM EDTA, no reactivation was observed.
Near-total restoration of QC activity after inactivation by
FIG. 4. Reactivation of human QC with monovalent and divalent metal ions. QC was nearly inactivated by the addition of 2 mM
dipicolinic acid in 50 mM Bis-Tris, pH 6.8. Subsequently, the enzyme
was subjected to dialysis against 50 mM Bis-Tris, pH 6.8, containing 1
mM EDTA. Reactivation of the enzyme was achieved by incubation of
the inactivated enzyme sample with metal ions at a concentration of 0.5
mM in the presence of 0.5 mM EDTA to avoid an unspecific reactivation
by traces of metal ions present in buffer solutions. Controls are given by
enzyme samples that were not inactivated but also dialyzed against the
EDTA solution as the inactivated enzyme (⫹EDTA) and by enzyme
samples that were dialyzed against buffer solutions without added
EDTA (⫺EDTA).
either dipicolinic acid or 1,10-phenanthroline was achieved by
incubating the protein for 10 min with 0.5 mM ZnSO4 in presence of 0.5 mM EDTA (Fig. 4). Partial restoration of QC activity
was similarly obtained using Co2⫹ and Mn2⫹ ions for reactivation. Even in the presence of 0.25 mM Zn2⫹, a reactivation to
25% of the original activity was possible (data not shown). No
reactivation was observed applying Ni2⫹, Ca2⫹, or K⫹ ions.
Similarly, incubation of fully active QC with these ions had no
effect on the enzyme activity.
DISCUSSION
After a more detailed comparison, human QC does not seem
to have much in common with its counterpart from the plant
Carica papaya except for the catalyzed reaction. In a recent
study of substrate specificity, we found a relatively similar
proficiency for glutaminyl cyclization by both enzymes (24).
However, differences were observed in binding and conversion
of peptides bearing the modified N-terminal glutaminyl residues ␥-hydrazide or ␥-methylamide. Although human QC is
inhibited by the hydrazide derivative (not papaya QC), only the
methylamide derivative is recognized and cyclized by the plant
enzyme. These results have already suggested differences in
the recognition modes of the substrate glutaminyl residue by
both enzymes. Additionally, we were unable to detect any inhibition of human QC by peptides containing N-terminal proline, which strongly inhibit papaya QC (13). Furthermore, in
striking contrast to the prominent inhibition of human QC by
imidazole derivatives, papaya QC was not influenced at all by
any of these compounds.
Similarly as with metal-dependent aminopeptidases, human
QC is inhibited by imidazole, 1,10-phenanthroline, and dipicolinic acid (17–19). In contrast to EDTA, these compounds all
show a planar structure, possibly enabling the interaction with
49778
Competitive Inhibition of Metal Enzyme Glutaminyl Cyclase
FIG. 5. Sequence alignment of human QC and other M28 family members of the metallopeptidase clan MH. Multiple sequence
alignment was performed using ClustalW at ch.EMBnet.org with default settings. The conservation of the zinc-ion ligating residues is shown for
human QC (hQC; GenBankTM number X71125), the zinc-dependent aminopeptidase from Streptomyces griseus (SGAP; Swiss-Prot number
P80561), and within the N-acetylated ␣-linked acidic dipeptidase (NAALADase I) domain (residues 274 –587) of the human glutamate carboxypeptidase II (hGCP II; Swiss-Prot number Q04609). The amino acids involved in metal binding are set in boldfaced type and underlined. In the case
of human QC, these residues are the putative counterparts to the peptidases. The shaded histidines (His-140 and His-330) indicate residues that
were identified as being essential for QC catalysis (9).
the active site-bound metal ion. Because of the complete reactivation by the addition of Zn2⫹ ions to apo QC, one can conclude that human and probably all mammalian QCs are Zn2⫹dependent. Recently, a relationship of the tertiary structure of
human QC and the aminopeptidase from Vibrio proteolyticus, a
prominent member of the clan MH family M28 of metallopeptidases, was proposed (9). Comparing the sequence of human
QC with those of two members of the clan MH (Fig. 5), the
binding motif His-Asp-Glu-Asp-His of the two Zn2⫹ ions present in this clan of hydrolases is also conserved in human QC.
Furthermore, as shown in another study (9), modification of
two of the identified histidine residues, His-140 and His-330,
which are probably necessary for metal binding (Fig. 5), leads
to a complete loss of catalytic activity. These data further
substantiate the fact that mammalian QCs evolved from an
ancestral metallo hydrolase and that at least one of the metal
binding sites is conserved. It remains unclear, however, how
the zinc ion(s) is involved in the catalysis of human QC. In the
metallo peptidases, the catalytic mechanism leads to the formation of a non-covalent tetrahedral intermediate after the
attack of a zinc-bound water molecule on the carbonyl group of
the scissile bond. Either in parallel or subsequently, zinc binding stabilizes the oxanion of the formed tetrahedral intermediate. Zinc ions increase the nucleophilicity of the peptide bond-
Competitive Inhibition of Metal Enzyme Glutaminyl Cyclase
attacking water molecule and polarize the scissile bond,
making it susceptible to nucleophilic attack during transition
state formation, with its progression to and the subsequent
collapse of the tetrahedral intermediate followed by amid bond
cleavage (20).
For the catalysis by human QC, the pH-dependence of substrate binding suggests that perhaps the metal ion could interact with the nitrogen of the N-terminal amino function of the
substrate. Because QC catalyzes an intramolecular cyclization,
the proper positioning of the substrate nitrogen in close proximity to the ␥-carbonyl carbon is probably of essential catalytic
importance. On the other hand, it seems likely that a metal ion
in the active site of QC acts by polarizing the ␥-amide group of
the substrate glutaminyl residue, simultaneously stabilizing
the oxanion formed by the nucleophilic attack of the ␣-nitrogen
on the scissile ␥-carbonyl carbon. Such an interaction of the
␥-carbonyl group of the substrate and the active-site metal ion
is also corroborated by the observed inhibition of human QC by
␥-hydrazide residues. Furthermore, the interaction of one active site zinc ion with the ␣-amino group of the substrate and
the polarization of the carbonyl group of the scissile peptide
bond by another are proposed steps in the catalysis of the
related aminopeptidase from V. proteolyticus (20). Accordingly,
the metal ion(s) of QC might serve as a binding site for the
imidazole-derived inhibitors and the substrate, with an unprotonated nitrogen interacting in analogy to the related
peptidase.
In contrast to the aminopeptidase from V. proteolyticus, however, increasing Zn2⫹ concentrations in QC assays (0.1 mM and
higher) considerably reduce QC activity, which also was observed in previous studies (3). Thus, it needs to be clarified
whether QC also possesses two metal ions bound to the apoenzyme or whether, during evolution from an aminopeptidase to
an acyltransferase, the already weak binding second Zn2⫹ ion
was lost (20) because a further Zn2⫹ ion was not needed for
exerting glutaminyl cyclization. The occupation of such a binding site at high concentrations of Zn2⫹ ions may block the
intramolecular reaction of the substrate.
It should be also noted, in this respect, that we could not
detect any proteolytic activity of QC. Moreover, 1-butaneboronic acid and peptide thiols (22, 23), potent inhibitors of
V. proteolyticus aminopeptidase, did not inhibit human QC,
supporting potential changes in the active site geometry of QC
compared with the aminopeptidases. Because there have been
extensive rearrangements during the evolution of the zinc hydrolase group, including remodeling of the active site upon
changes in zinc ligation (20, 21), only the solution of the protein
structure will finally clarify the binding modes of substrate
and inhibitor.
In contrast to its mammalian counterparts, papaya QC is not
inhibited by metal chelators, suggesting a metal-independent
mechanism. However, for the cyclization reaction, the nitrogen
of the ␣-amino group of the glutaminyl residue also needs to be
deprotonated (Scheme 1), and both enzymes show a similar
catalytic proficiency of catalysis. How the same catalytic reac-
49779
tion of such structurally divergent protein catalysts is maintained will remain obscure until the solution of the threedimensional structures of both proteins.
In summary, we present here the first systematic structureactivity study of inhibitors for a mammalian QC. Because there
is no reliable active site model for any QC available to date,
there was only minimal information to identify the structural
features that need to be incorporated into potent QC inhibitors.
Besides the identification of N-1 imidazole derivatives as
highly potent competitive inhibitors, the results revealed
human QC as a metal-dependent enzyme as shown by the
following: (a) the pH-dependence of inhibition by imidazole and
imidazole derivatives; (b) the inactivation of QC by the metalchelating reagents 1,10-phenanthroline and dipicolinic acid; (c)
the reactivation of the QC-apoenzyme by bivalent metal ions;
and (d) the conservation of metal-binding residues in the primary structure of QC. Finally, the observed impact of structural modifications of the imidazole derivatives on their QCinhibitory potency can serve as a starting point for further,
rationally driven inhibitor designs.
Acknowledgments—The technical assistance of J. Zwanzig and H.
Cynis is gratefully acknowledged. We are grateful to Drs. S.A. Hinke
and J.A. Pospisilik for critical comments on the manuscript.
REFERENCES
1. Richter, K., Kawashima, E., Egger, R., and Kreil, G. (1984) EMBO J. 3,
617– 621
2. Messer, M. (1963) Nature 197, 1299
3. Busby, W. H., Jr., Quackenbush, G. E., Humm, J., Youngblood, W. W., and
Kizer, J. S. (1987) J. Biol. Chem. 262, 8532– 8536
4. Dahl, S. W., Slaughter, C., Lauritzen, C., Bateman, R. C. Jr., Connerton, I.,
and Pedersen, J. (2000) Protein Expr. Purif. 20, 27–36
5. Fischer, W. H., and Spiess, J. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,
3628 –3632
6. Pohl, T., Zimmer, M., Mugele, K., and Spiess, J. (1991) Proc. Natl. Acad. Sci.
U. S. A. 88, 10059 –10063
7. Zerhouni, S., Amrani, A., Nijs, M., Smolders, N., Azarkan, M., Vincentelli, J.,
and Looze, Y. (1998) Biochim. Biophys. Acta 1387, 275–290
8. Oberg, K. A., Ruysschaert, J. M., Azarkan, M., Smolders, N., Zerhouni, S.,
Wintjens, R., Amrani, A., and Looze, Y. (1998) Eur. J. Biochem. 258,
214 –222
9. Bateman, R. C., Jr., Temple, J. S., Misquitta, S. A., and Booth, R. E. (2001)
Biochemistry 40, 11246 –11250
10. Schilling, S., Hoffmann, T., Rosche, F., Manhart, S., Wasternack, C., and
Demuth, H.-U. (2002) Biochemistry 41, 10849 –10857
11. Gololobov, M. Y., Song, I., Wang, W., and Bateman, R. C., Jr. (1994) Arch.
Biochem. Biophys. 309, 300 –307
12. Temple, J. S., Song, I., Burns K. H., and Bateman, R. C., Jr. (1998) Korean
J. Biol. Sci. 2, 243–248
13. Gololobov, M. Y., Wang, W., and Bateman, R. C., Jr. (1996) Biol. Chem.
Hoppe-Seyler 377, 395–398
14. Schilling, S., Hoffmann, T., Wermann, M., Heiser, U., Wasternack, C., and
Demuth, H.-U. (2002) Anal. Biochem. 303, 49 –56
15. Tsuru, D., Fujiwara, K., and Kado, K. (1978) J. Biochem. (Tokyo) 84, 467– 476
16. Ellis, K. J., and Morrison, J. F. (1982) Methods Enzymol. 87, 405– 426
17. Czekay, G., and Bauer, K. (1993) Biochem. J. 290, 921–926
18. Kelly, J. A., Slator, G. R., Tipton, K. F., Williams, C. H., and Bauer, K. (2000)
J. Biol. Chem. 275, 16746 –16751
19. Aoyagi, T., Yoshida, S., Matsuda, N., Ikeda, T., Hamada, M., and Takeuchi, T.
(1991) J. Antibiot. (Tokyo) 44, 573–578
20. Lowther, W. T., and Matthews, B. W. (2002) Chem. Rev. 102, 4581– 4607
21. Wouters, M. A., and Husain, A. (2001) J. Mol. Biol. 314, 1191–1207
22. Baker, J. O., and Prescott, J. M. (1983) Biochemistry 22, 5322–5331
23. Huntington, K. M., Bienvenue, D. L., Wei, Y., Bennett, B., Holz, R. C., and Pei,
D. (1999) Biochemistry 38, 15587–15596
24. Schilling, S., Manhart, S., Hoffmann, T., Ludwig, H.-H., Wasternack, C., and
Demuth, H.-U. (2003) Biol. Chem. 384, in press
Continuous Assays of Glutaminyl Cyclase: From Development to Application
Stephan Schilling and Hans-Ulrich Demuth*
Probiodrug AG, Weinbergweg 22, 06120 Halle/Saale, Germany
∗
To whom correspondence should be addressed at probiodrug AG, Weinbergweg 22, 06120 Halle,
Germany
tel.: 49 345 5559900 fax: 49 345 5559901 email: [email protected]
Abstract
Glutaminyl cyclase (QC, EC 2.3.2.5) catalyses the formation of pyroglutamyl residues from glutamine at the Nterminus of peptides and proteins. In previously applied assays, QC activity was determined by either analysing the
products formed using HPLC coupled with photometric or fluorometric detection, radioimmunoassay, or by detecting
the release of ammonia spectrophotometrically. Although these methods are sensitive, they are all discontinuous and
therefore time-consuming and laborious. To conduct detailed kinetic investigation of QC-catalysis, we developed
coupled continuous assays suitable for microplates which allow now convenient determination of QC activity. The
methods either use pyroglutamyl aminopeptidase or glutamate dehydrogenase as auxiliary enzymes, which results in the
liberation of chromophores or fluorophores such as pNA, AMC, βNA or in the conversion of the chromophore
NADH/H+ into NAD+, respectively.
The assays were applied in various enzyme isolation and characterization studies, using crude protein solutions as well
as purified enzyme in pH-dependence, substrate and inhibitor specificity investigations. Depending on the respective
analytical task, both assays complement each other. Therefore, different enzymatic properties could be explored in more
detail. Since the employed strategy of assay development could be of interest also for the analysis of other enzymes, the
methods are described here in a comprehensive manner.
Introduction
Several peptides and proteins contain pyroglutamic acid at their N-terminus. Initially, pyroglutamyl
formation was assumed to result from a spontaneous cyclisation reaction of a N-terminal glutamine
or glutamic acid residue. However, specific enzymatic conversion of glutamine by glutaminyl
cyclase (EC 2.3.2.5) has been discovered in plant and animal tissues [1,2,3,4].
Although a QC was first explored in papaya latex, its physiological function in the plant is still
enigmatic. More recently, however, QCs were also identified in several other plant species,
suggesting a general physiological significance of this protein [5]. In contrast to plant QC, several
physiological substrates and products of mammalian QC-activities could be identified.
Pyroglutamic acid is present, for instance in the hormones Thyrotropin releasing hormone (TRH),
gonadoliberin (GnRH) and gastrin, neurotensin and chemokines of the monocyte chemotactic
protein (MCP) family. The formation of this N-terminal 5-oxoproline residue has shown to cause
the bioactive structure of the hormones and to improve their stability towards N-terminal
proteolysis [6,7,8]. Interestingly, plant and animal QCs seem to be very similar at a first glance.
Both enzyme forms are expressed via the secretory pathway, carry carbohydrates and are
monomeric proteins with similar molecular masses of 33-40 kDa [5,9,10]. Furthermore, all QCs are
strictly specific for L-glutamine in the N-terminal position of the substrates and their kinetic
behaviour was found to obey the Michaelis-Menten equation [10,11,12]. The primary and
secondary structures of the QCs from C. papaya and that of the highly conserved QC from
mammals, however, did not reveal homology [5,13,14]. Due to this apparent divergence, a detailed
comparison of the catalytic properties of the QC forms could be helpful for deepening the
understanding of pyroglutamyl formation and to identify, whether the different QC forms catalyse
the pyroglutamyl formation by the same mechanism.
A detailed enzymatic characterisation of QC catalysis and inhibition, however, was hampered by
the lack of handy assays. In previously applied methods, QC activity was determined by either
analysing the products formed using HPLC linked to photometric or fluorimetric detection [11,10]
or radioimmunoassay [3,15] using antibodies directed against TRH, or by detecting the release of
ammonia spectrophotometrically in a coupled enzymatic assay [16]. Although the methods are
sensitive, they are all discontinuous and therefore time-consuming and laborious. Furthermore,
some of these methods can be only applied for one certain substrate, thus hampering detailed
substrate specificity studies. In another approach, the change in absorbance occurring due to the
formation of an intramolecular amid bond during N-terminal pyroglutamyl formation by QC is
detected [17]. Although this assay allows a continuous data monitoring, the observed changes in
absorbance are very small, making the assay insensitive. Furthermore, due to measuring the
absorbance change at 220 nm, at the wavelength characteristic for the n→π* electron transition of
peptide bonds, enzyme activity in crude samples cannot be determined because of the huge
background. Accordingly, also the high initial absorbance of large peptides hinders the
determination of catalytic parameters for such QC substrates.
Due to these disadvantages, the development of new assays was triggered that allow a) the
convenient and fast determination of QC activity, making it suitable during protein purification and
characterization and b) to easily determine the specificity of QCs for an assortment of substrates of
different size and structure. This flexibility was obtained by developing coupled continuous assays
that utilize different auxiliary enzymes. The inability of detecting the intramolecular amid
transferase reaction in buffered systems could be compensated by well-observable coupled
reactions, enabling detailed QC-characterisation studies such as substrate and inhibitor specificity,
influence of ionic strength and pH-dependence of the kinetic parameters.
Body of text
Coupled enzymatic assays - theoretical considerations. Enzyme catalysed reactions are most often
analysed using spectrophotometric or fluorimetric detection, since the detectors, i.e. the
photometers or fluorimeters, are relatively inexpensive and present in nearly all life science
laboratories. However, many enzyme catalysed reactions cannot be monitored directly, since
substrate conversion does not result in a change of the absorbance or fluorescence characteristics, as
for instance the case in kinase-, phosphatase- or many transferase-catalysed reactions. Therefore,
coupled enzymatic assays were established, using an auxiliary reaction that results in a change in
absorbance or fluorescence. In the coupled reaction, one of the products of the reaction that should
be analysed is consumed. The simplest case, which is also valid for the assays described below, can
be represented schematically
A
k1
primary enzyme
B
k2
C
auxiliary enzyme
For a reliable assay, the following assumptions have to be fulfilled [18]: a) k1 represents a zeroorder rate constant, i.e. the concentration of A does practically not change during the observed
reaction time, b) the second reaction is irreversible, and c) k2 is a rate constant of first order, which
requires that the concentration of B is always much lower than the Michaelis constant of the
auxiliary enzyme for B ([B]<< KB). Based on these assumptions, the rate equation focussed on
formation and consumption of B, which is the prerequisite of the observed spectroscopic changes, is
d[B]
= k1 – k2[B]
dt
(1)
which provides after integration
[B]=
k1
(1 – e - k2t )
k2
(2)
As can be seen from this equation, if time runs to infinity (t→∞), a constant concentration of B is
reached, the so-called steady state concentration [Bss], characterised by a linear progress of the
formation of C. In a practical view, progress curves are usually indistinguishable from linearity, if
95 % of the intermediate steady-state concentration is reached, which is sufficient for providing
reliable results. The time to reach this state is characterised by a “lag phase”, the progress curve
shows an exponential increase. After rearrangement of equation 2, substitution of –Vmax2 t/KB for
-k2 t and [Bss] for (k1/k2), the following equation is obtained, which offers the opportunity to
calculate the time until reaching assay conditions that provide progress curves indistinguishable
from linearity
Vmax2 =
- KB ln(1 – [B]/[Bss])
t*
.
(3)
In this equation, t* denotes the time to reach a certain fraction of [Bss] (i.e., [B]/[Bss]), which is
dependent on the concentration and specificity of the auxiliary enzyme for its substrate, indicated
by Vmax2/KB. Therefore, with knowledge of the specificity of the respective auxiliary enzyme, one
can calculate the amount of protein required to obtain a reliable assay, without consuming excessive
protein quantities. As follows, equation 3 was applied for the development of two different
continuous assays for determination of QC activity.
Coupling QC to Pyroglutamyl Aminopeptidase (pGAP) catalysis - development. Coupling the
cyclising activity of QC to a peptidase was accomplished by use of dipeptide surrogates that are
prone to cleavage after conversion by QC. Accordingly, potential assay substrates possessing N-
terminal glutamine are Gln-pNA, Gln-βNA or Gln-AMC. After cyclisation by QC, the respective
intermediates pGlu-pNA, pGlu-βNA or pGlu-AMC are hydrolysed by pGAP, liberating a
chromophoric or fluorogenic group. Since the spectrophores are liberated in equimolar amounts to
the glutaminyl-substrate converted, QC-activity can be calculated from standard curves. The
reactions are exemplified for the turnover of Gln-βNA in Figure 1.
For assay development, the bacterial pyroglutamyl aminopeptidase from Bacillus amyloliquefaciens
was chosen. This well-characterised cysteine protease shows a broad substrate specificity, suitable
stability and is commercially available. With regard to specificity, the potential intermediates in QC
assay have shown to be among the best substrates of this enzyme [19,20,21,22]. The time to reach
virtual steady state conditions in the QC assay, if 1 U/ml auxiliary enzyme is applied, were
calculated according to equation 3 using the available specificity data of pGAP (table 1). Due to the
relatively low specificity of pGAP towards the intermediate pGlu-pNA, the time until observation
of linear progress curves is approximately 2 min for the chromophoric substrate Gln-pNA. In
contrast, when using Gln-βNA as QC substrate, virtual steady state conditions are observed within
one second after initiation of the reaction caused by the high specificity of pGAP for the
intermediate pGlu-βNA.
In fact, linear progress curves were observed according to such calculation. They are exemplified
for Gln-βNA and Gln-AMC in Figure 2. For all substrates, there was a linear relationship between
the QC concentration and the observed rate, indicating the linear dependence of the assay on
conversion of the QC substrate (not shown). Finally, the assay could be applied for recombinant
human or mouse and purified papaya QC. The now possible characterisation studies enabled the
comparison of the QC forms concerning differences and similarities of their catalysis. Due to the
shorter lag times observed with the fluorogenic substrates (Table 1), the assays using Gln-AMC or
Gln-βNA provide a higher flexibility. Small alterations in the activity of the auxiliary enzyme do
not affect the assay, because the auxiliary enzyme activity is still excessive to provide reliable
results, i.e. the “lag times” are always shorter than the time required for starting the reaction and
mixing of the samples. Therefore, most of the characterising studies shown below were carried out
using the fluorogenic substrates.
Coupling QC to Pyroglutamyl Aminopeptidase (pGAP) catalysis – application. Applications of
spectroscopic enzyme assays range from the identification of enzyme activity in tissues,
quantification of enzymatic activity during protein purification, protein characterisation in terms of
pH-dependence of catalysis, substrate specificity and for inhibitor screening. Recently, we applied
the assay during purification of papaya QC [23]. Although the continuous data monitoring already
accelerated the enzyme determination during the purification procedure, its application is much
more important during characterisation studies, since many assay reactions have to be performed,
thus favouring the continuous assays. Moreover, only the analysis of kinetic parameters
investigating a wide substrate concentration range makes it possible to detect differences in kinetic
mechanisms or models.
Hence, the plots of substrate concentration versus the respective velocities obtained for Gln-AMC
and Gln-βNA follow different kinetic laws. Whereas the kinetic data for Gln-AMC readily
resembled Michaelis-Menten kinetics in the concentration range limited by substrate solubility,
Gln-βNA showed discernible substrate inhibition (Figure 3). Interestingly, papaya QC showed a
higher selectivity for Gln-βNA, but was inhibited by the substrate with similar potency. To our
knowledge, Gln-βNA is the only QC substrate showing differences to Michaelis-Menten-kinetics,
which could be indicative for a similar catalytic action of both enzymes.
Subsequently, the pH-dependence of the catalytic parameters kcat and KM for conversion of GlnAMC by human QC was investigated. The pH-range used was restricted to 5.5-8.5, due to the
limited stability of QC and the auxiliary enzyme at more basic and acidic pH. Similar experiments
were already performed with papaya QC using Gln-tBE as substrate [17], revealing that the
catalytic activity depends only on changes of substrate binding. Apparently, the substrate having a
protonated amino group was not bound by the active site. The rate constant kcat did not change in
the investigated pH-range. Also human QC-catalysis exhibited only a dependence in terms of
substrate binding, reflected by a pH-dependent change of KM, and a pH-independent kcat (Figure 4).
Fitting of the pH-dependent kinetic data of KM to an equation that accounts for two dissociating
groups revealed a pKa-value that is very close to the pKa of the substrate amino group and a second
pKa, probably representing a dissociating group of the enzyme. Thus, human and papaya QC bind
only N-terminally unprotonated substrate molecules in a catalytically productive manner, indicating
some general similarity in the catalytic mechanism, in spite of a lack of structural homology.
Differences, however, where observed in the binding of inhibitory compounds. Whereas papaya QC
was inhibited by peptides bearing N-terminal proline [12], human QC was not. We found, however,
competitive inhibition of human QC by peptides bearing N-terminal γ-glutamyl-hydrazide residues
(Figure 5). Furthermore, human QC was also inhibited by imidazole derivatives which contrasts
with plant QC (not shown). These results suggest differences of plant and human QC concerning
substrate conversion, apparently due to differences in the substrate and inhibitor recognition modes.
Although the described assays have shown to be suitable for many applications, there is one major
disadvantage. Only QC substrates can be used, whose conversion yields finally chromophoric or
fluorogenic groups. Thus, the substrate spectrum is limited to variations of the chromophores and
fluorophores, which hampers a detailed substrate specificity investigation. Therefore, alternative
assays were needed, that overcome this drawback without waiving a continuous data monitoring.
Coupling QC to glutamic dehydrogenase (GDH) catalysis – development and application of an
alternative assay. This assay is based on quantification of ammonia, that is liberated by cyclisation
of glutamine. The auxiliary enzyme in this assay is glutamic dehydrogenase, converting ammonia,
α-ketoglutaric acid and NADH/H+ into glutamic acid and NAD+. Since the absorbance
characteristics of NADH/H+ changes by oxidation, its conversion can be followed at 340 nm. QC
activity can be subsequently quantified by calculation of the liberated ammonia from standard
curves of ammonia under assay conditions. The assay reactions are illustrated in Figure 6.
Originally, the assay was developed as a discontinuous method [16], probably due to the relatively
low affinity of GDH towards ammonia. In turn, this low reactivity leads to very high auxiliary
enzyme concentrations necessary to implement conditions that enable a continuous data monitoring
according to equation 3. The usage of common cuvettes for spectrophotometers requiring sample
volumes of 1-2 ml, however, causes the consumption of tremendous amounts of GDH, making the
assay prohibitive. A calculation of the required auxiliary enzyme amount (equation 3) resulted in 30
U/ml GDH needed to reach a virtual steady state 20 s after initiation of the reaction (assuming a KM
of 3.2 mM of GDH for ammonia [24] and that one Unit of GDH refers to saturating concentrations
of all substrates). Due to implementation of microplate readers for assay development, it was
possible to reduce the assay volume to 250 µl, thus keeping the required auxiliary enzyme amount
low, but still providing convenient volumes for pipetting. Furthermore, the use of the microplates,
that were already applied in the fluorometric assays, enable a fast determination of QC activity in
many samples at the same time, thus accelerating the determinations enormously. Finally, due to
the detection of ammonia that is liberated from the QC-substrates the assay can be implemented for
a fast examination of a variety of glutaminyl-peptides.
In fact, linear progress curves were obtained according to the predicted conditions, and most
importantly there was a linear relationship between QC-concentration and initial velocity, indicating
that the assay provides reliable results (Figure 7). Subsequently, a detailed substrate specificity
investigation was performed using about 40 newly synthesised substrates (Schilling et al.,
submitted), showing that the assay is applicable independently from changes of substrate amino
acid composition and peptide size. Moreover, since the auxiliary enzyme was not influenced
significantly by potassium chloride concentrations up to 0.5 M, the method was also implemented
to investigate the influence of ionic strength on conversion of different substrates. Interestingly, the
observed changes of the specificity constant kcat/KM were substrate dependent (Figure 8). Neither
papaya QC, nor human QC revealed altered activity by increasing ionic strength towards peptides
with uncharged backbone and residues. Both enzymes, however, displayed a significant increase in
activity towards peptides comprising positively charged amino acid residues. Thus, besides the
similar pH-dependence of catalysis for both enzymes, also the behaviour of the different QC-forms
in environments with differing dielectric constants is similar, which could be also indicative for
analogous catalytic mechanisms. Finally, this conclusion was also corroborated by the ability of
both, human and papaya QC, to cyclise N-terminal β-homoglutaminyl residues with the same
catalytic efficiency (not shown).
Conclusions
The application of theoretical deductions [18] facilitated the development of the first coupled
enzymatic assays for Glutaminyl cyclase (QC) activity. Due to the use of different enzymatic
reactions for coupling to QC catalysis, many characterisation studies could be performed including
pH-dependence, inhibitory and substrate specificity. In this regard, the general differences in the
coupling strategy, i.e. the consumption of the QC-products, either the pyroglutamyl peptide by
pGAP or the liberated ammonia by GDH, led to a compensation of respective disadvantages of both
assays. For instance, traces of ammonia in a sample hamper the QC-determination in the GDHcoupled assay, but show no effect in the assays using pGAP, resulting in the preferred usage of the
latter enzyme for assays during enzyme purification. When investigating different peptide
substrates, however, only the coupling to ammonia production provided satisfying results since a
large substrate spectrum that can be investigated. Thus, the use of the different assay coupling
strategies enabled the convenient determination of QC activity in different fields of protein
characterisation.
Finally, the demonstrated strategy to develop continuous assay techniques could also be used to
modify discontinuous assays for other enzymes or to develop new ways for their catalytic
characterisation by implementing different coupling strategies.
Acknowledgement
We thank K. Zunkel, H. Cynis and J. Zwanzig for valuable technical assistance.
References
[1] M. Messer, Enzymatic cyclization of L-glutamine and L-glutaminyl peptides, Nature 4874
(1963), 1299.
[2] M. Messer and M. Ottesen, Isolation and properties of glutamine cyclotransferase of dried
papaya latex, Biochim Biophys Acta 92 (1964), 409-411.
[3] W.H.J. Busby, G.E. Quackenbush, J. Humm, W.W. Youngblood, and J.S. Kizer, An
enzyme(s) that converts glutaminyl-peptides into pyroglutamyl-peptides. Presence in
pituitary, brain, adrenal medulla, and lymphocytes, J Biol Chem 262 (1987), 85328536.
[4] W.H. Fischer and J. Spiess, Identification of a mammalian glutaminyl cyclase converting
glutaminyl into pyroglutamyl peptides, Proc Natl Acad Sci U S A 84 (1987), 36283632.
[5] S.W. Dahl, C. Slaughter, C. Lauritzen, R.C.Jr. Bateman, I. Connerton, and J. Pedersen, Carica
papaya glutamine cyclotransferase belongs to a novel plant enzyme subfamily: cloning
and characterization of the recombinant enzyme, Protein Expr Purif 20 (2000), 27-36.
[6] E. Van Coillie, P. Proost, I. Van Aelst, S. Struyf, M. Polfliet, I. De Meester, D.J. Harvey, J.
Van Damme, and G. Opdenakker, Functional comparison of two human monocyte
chemotactic protein-2 isoforms, role of the amino-terminal pyroglutamic acid and
processing by CD26/dipeptidyl peptidase IV, Biochemistry 37 (1998), 12672-12680.
[7] G.N. Abraham and D.N. Podell, Pyroglutamic acid. Non-metabolic formation, function in
proteins and peptides, and characteristics of the enzymes effecting its removal, Mol
Cell Biochem 38 Spec No (1981), 181-190.
[8] S.A. Hinke, J.A. Pospisilik, H.-U. Demuth, S. Manhart, K. Kühn-Wache, T. Hoffmann, E.
Nishimura, R.A. Pedersen, and C.H.S. McIntosh, Dipeptidyl peptidase IV
(DPIV/CD26) degradation of Glucagon, J Biol Chem 275 (2000), 3827-3834.
[9] S. Zerhouni, A. Amrani, M. Nijs, N. Smolders, M. Azarkan, J. Vincentelli, and Y. Looze,
Purification and characterization of papaya glutamine cyclotransferase, a plant enzyme
highly resistant to chemical, acid and thermal denaturation, Biochim Biophys Acta
1387 (1998), 275-290.
[10] T. Pohl, M. Zimmer, K. Mugele, and J. Spiess, Primary structure and functional expression of
a glutaminyl cyclase, Proc Natl Acad Sci U S A 88 (1991), 10059-10063.
[11] A.P. Consalvo, S.D. Young, B.N. Jones, and P.P. Tamburini, A rapid fluorometric assay for
N-terminal glutaminyl cyclase activity using high-performance liquid chromatography,
Anal Biochem 175 (1988), 131-138.
[12] M.Y. Gololobov, W. Wang, and R.C.J. Bateman, Substrate and inhibitor specificity of
glutamine cyclotransferase (QC), Biol Chem Hoppe Seyler 377 (1996), 395-398.
[13] K.A. Oberg, J.M. Ruysschaert, M. Azarkan, N. Smolders, S. Zerhouni, R. Wintjens, A.
Amrani, and Y. Looze, Papaya glutamine cyclase, a plant enzyme highly resistant to
proteolysis, adopts an all-beta conformation, Eur J Biochem 258 (1998), 214-222.
[14] S. Schilling, T. Hoffmann, F. Rosche, S. Manhart, C. Wasternack, and H.U. Demuth,
Heterologous expression and characterization of human glutaminyl cyclase: evidence
for a disulfide bond with importance for catalytic activity, Biochemistry 41 (2002),
10849-10857.
[15] J.B. Koger, J. Humm, and J.S. Kizer, Assay of glutaminylpeptide cyclase, Methods Enzymol
168 (1989), 358-365.
[16] R.C.J. Bateman, A spectrophotometric assay for glutaminyl-peptide cyclizing enzymes, J
Neurosci Methods 30 (1989), 23-28.
[17] M.Y. Gololobov, I. Song, W. Wang, and R.C.J. Bateman, Steady-state kinetics of glutamine
cyclotransferase, Arch Biochem Biophys 309 (1994), 300-307.
[18] W.R. McClure, A kinetic analysis of coupled enzyme assays, Biochemistry 8 (1969), 27822786.
[19] K. Fujiwara, R. Kobayashi, and D. Tsuru, The substrate specificity of pyrrolidone carboxylyl
peptidase from Bacillus amyloliquefaciens, Biochim Biophys Acta 570 (1979), 140148.
[20] K. Fujiwara and D. Tsuru, New chromogenic and fluorogenic substrates for pyrrolidonyl
peptidase, J Biochem (Tokyo) 83 (1978), 1145-1149.
[21] D. Tsuru, K. Fujiwara, and K. Kado, Purification and characterization of Lpyrrolidonecarboxylate peptidase from Bacillus amyloiliquefaciens, J Biochem
(Tokyo) 84 (1978), 467-476.
[22] P.M. Cummins and B. O'Connor, Pyroglutamyl peptidase: an overview of the three known
enzymatic forms, Biochim Biophys Acta 1429 (1998), 1-17.
[23] S. Schilling, T. Hoffmann, M. Wermann, U. Heiser, C. Wasternack, and H.-U. Demuth,
Continuous spectrometric assays for glutaminyl cyclase activity, Anal Biochem 303
(2002), 49-56.
[24] C. Frieden, Glutamic Dehydrogenase, 234 (1959), 2891.
[25] K.J. Ellis and J.F. Morrison, Buffers of constant ionic strength for studying pH-dependent
processes, Methods Enzymol. 87 (1982), 405-426.
Tables
Table 1: Times to reach virtual steady state conditions in QC assay by coupling to
pyroglutamyl aminopeptidase (1 U/ml), calculated according to equation 3. The
kinetic data for pGAP were obtained from references [20] and [21]a.
time to reach
95 % [B,ss] in QC
assay at 1 U/ml
Substrate
KB (mM)
Vmax2
(µmol mg-1 min-1)
pGlu-pNA
0.69
0.56
2 min
pGlu-AMC
0.33
5.7
6s
pGlu-βNA
0.13
20.0
<1s
Figure captions
Figure 1: Representation of the coupled QC-assay using Gln-βNA as substrate. In the initial
reaction, Gln-βNA is converted by QC into pGlu-βNA. Subsequently, the intermediate is
cleaved by the abundant pyroglutamyl aminopeptidase into pyroglutamic acid and the
fluorophore 2-naphthylamine, resulting in an increase in the observed fluorescence.
Figure 2: Progress curves of the conversion of Gln-βNA and Gln-AMC by QC, investigated
by coupling the reaction to pGAP catalysis. According to a calculation (Equation
3), linear progress curves were observed directly after initiation of the reaction
(Gln-βNA and Gln-AMC). Assays were carried out in 0.05 M Tris/HCl, pH 8.0 at
30 °C. The substrate and QC concentrations were 0.25 mM and 0.9 nM,
respectively.
Figure 3: Dependence of the conversion-rate of Gln-βNA from the substrate concentration,
determined for human (α) and papaya QC (Β). The resulting graphs were obtained by
fitting the data to the general equation of substrate inhibition. Human QC (KM= 70
±3 µM, kcat= 21 ±1 s-1, Ki= 1.21 ±0.07 mM) showed a reduced specificity compared
to papaya QC (KM= 36 ±2 µM, kcat= 49 ±1 s-1, Ki= 1.14 ±0.05 mM), but the
catalysis was inhibited to similar extends by the substrate. Reactions were carried
out in 0.05 M Tris/HCl, pH 8.0 (human QC) or 0.05 M Tricine/NaOH, pH 8.0
(papaya QC) at 30 °C.
a
The unit definition refers to pGlu-pNA. One unit of pGAP is defined as 1 µmol substrate converted per min under the
described conditions.
Figure 4: The pH-dependence of Gln-AMC conversion by human QC. At varying pH-values,
the kinetic parameters KM and kcat were determined, and the logarithmic values
plotted. Whereas the kcat-value (Β) was independent from pH, the KM-values (α)
increased in the acidic and basic pH-region. Fitting the data to an equation that accounts
for two dissociating groups resulted in pKa-values of 6.81 ±0.04 and 8.6 ±0.1. The
former value is in good agreement with the pKa of the substrate whereas the latter
probably reflects a dissociating group of the enzyme. Reactions were carried out in
a buffer system providing a constant ionic strength over the entire pH-range,
consisting of 0.06 M acetic acid, 0.06 M Mes and 0.12 M Tris [25] at 30 °C.
Figure 5: The array of curves obtained for the conversion of Gln-AMC by human QC in
presence of varying concentrations of the inhibitory active peptide H-Glu(NHNH2)-Ser-Pro-Thr-Ala-NH2. Data points were fitted according to the general
equation for competitive inhibition. The resulting Ki-value was 0.697 ±0.003
mM. The assay was carried out in 0.05 M Tris/HCl, pH 8.0, containing 5 mM
EDTA. The substrate concentrations ranged from 1 mM to 0.125 mM.
Figure 6: Representation of the QC-assay using GDH as auxiliary enzyme and an N-terminal
Glutaminyl peptide as substrate. In the initial reaction, the respective pyroglutamyl
peptide and ammonia are formed. Subsequently, ammonia, α-ketoglutaric acid and
NADH/H+ are converted into glutamic acid and NAD+ catalysed by GDH. The
consumption of NADH/H+ can be observed at 340 nm.
Figure 7: Linear dependence of the initial rate of conversion on concentration of human QC
using GDH as auxiliary enzyme. The inset shows two progress curves, in the
sample containing human QC (12 nM), a linear decrease of
absorbance was observed. Without added QC, the decrease in absorbance was negligible.
Reactions were carried out in 0.05 M Tris/HCl pH 8.0, containing 5 mM EDTA.
Figure 8: The influence of ionic strength on the specificity constant kcat/KM for conversion of
various substrates by human and papaya QC. For most peptides, there was little
effect of changes in ionic strength detected. However, human and papaya QC
specificity towards positively charged peptides increased significantly by addition of
0.5 M KCl. Without additional salt added, the ionic strength was 0.029 M, corresponding
to a 0.05 M Tris- or Tricine buffer.
Figure 1:
O
H2N
O
O
H2N
QC
HN
O
O
N
H
NH
pGAP
O
N
H
OH
+
NH3
NH2
Figure 2:
Fluorophore concentration (µM)
8
7
6
5
4
3
2
1
0
0
200
400
Time (s)
600
800
Rate (mM min-1 mg-1)
Figure 3:
250
200
150
100
50
0
0
0,5
1
1,5
[Gln-ßNA] in mM
2
Figure 4:
1,6
0,8
1,4
0,6
0,4
1
0,2
0,8
0
0,6
-0,2
0,4
0,2
-0,4
0
-0,6
6
7
pH
8
lg kcat
pKM
1,2
Figure 5:
Rate (RFU/min)
250
200
150
100
50
0
0
0,2
0,4
0,6
0,8
[Gln-AMC] in mM
1
Figure 6:
peptide
peptide
HN
O
H2N
O
NH
QC
+
NH3
NH
O
NH2
O
GDH
+
NH3 + NADH/H + α−ketoglutaric acid
NAD+ + glutamic acid
Absorbance (340 nm)
QC activity (µM/min)
Figure 7:
30
25
20
15
1,2
1
0,8
0,6
0,4
0,2
0
0
250
500
750
Time (s)
10
5
0
0
2
4
6
QC (nM)
8
10
12
Figure 8:
1400
Papaya QC,
0.5M KCl
1200
1000
800
600
400
200
0
1800
1600
1400
Human QC
Human QC,
0.5M KCl
1200
1000
800
600
400
200
0
Q
-N
H
Q 2
-ß
N
Q A
W
A
Q
A
Q
E
Q
RG
Q I
K
R
Q L
ED
L
1600
Papaya QC
Specificity (mM-1s-1)
1800
2000
Q
-N
H
Q 2
-ß
N
Q A
W
A
Q
A
Q
E
Q
RG
Q I
K
R
Q L
ED
L
Specificity (mM-1s-1)
2000
Inhibition of glutaminyl cyclase prevents the formation of
pGlu3-amyloid-β related peptides
Stephan Schilling, Torsten Hoffmann, Susanne Manhart, Matthias Hoffmann, Hans-Ulrich
Demuth∗
Probiodrug AG, Weinbergweg 22, 06120 Halle/Saale, Germany
Running Title: Discovering the glutamate cyclase activity of glutaminyl cyclase
∗
To whom correspondence should be addressed at probiodrug AG, Weinbergweg 22, 06120
Halle, Germany tel.: +49 345 5559900 fax: +49 345 5559901
email: [email protected]
Abbreviations: AD, Alzheimer`s disease; APP, amyloid precursor protein; BACE, βNA, 2naphthylamine; pGlu, pyroglutamic acid; QC, glutaminyl cyclase; EC, glutamyl cyclase; DP
IV, dipeptidyl peptidase IV;
Aβ(3-11)a, Amyloid β-peptide 3-11 amide, GnRH,
Gonadotropin releasing-hormone; TRH, thyrotropin releasing-hormone
Abstract
N-terminal pyroglutamate (pGlu) has been identified as one of the major amyloid-β (Aβ)
peptide components of plaques found in the brains of Alzheimer’s disease (AD) patients.
Several related Aβ peptides contain N-terminal glutamate residues at positions 3 and 11. In
parallel or subsequent to posttranslational β- and γ-secretase cleavage of the Aβ precursor
protein (APP), N-terminal processing generates pGlu-Aβ peptides, molecular species are
much more resistant to further degradation by aminopeptidases. Such N-terminal pGlupeptide formation can be facilitated by glutaminyl cyclase (QC), a metal-dependent enzyme
abundant in the brain and thought to be responsible for the ultimate processing of bioactive
neuropeptides such as TRH and neurotensin during their maturation in the secretory pathway.
To clarify whether QC can also recognize Aβ-related peptides, the turnover of Gln3-Aβ(111)a, Aβ(3-11)a, Gln3-Aβ(3-11)a, Aβ(3-21)a, Gln3-Aβ(3-21)a and Gln3-Aβ(3-40) by the
enzyme was investigated. Unexpectedly, we found that recombinant human QC as well as
QC-activity from procine brain extracts catalyze both the N-terminal glutaminyl as well as
glutamate cyclization. Most striking was the finding that cyclase-catalyzed Glu1-conversion is
favored around pH 6.0 while Gln1-conversion to pGlu-derivatives occurs with a pH-optimum
of around 8.0. Since the formation of pGlu-Aβ-related peptides can be suppressed by
inhibition of recombinant human QC and QC-activity from porcine pituitary extracts, the
enzyme QC is suggested as a novel target in drug development for AD-treatment.
Introduction
Regulatory peptides such as GnRH, TRH and neurotensin, and the cytokines MCP-1 through
4, require N-terminal pyroglutamate in order to exert their respective biological functions
(1;2). Early studies have suggested that the formation of pyroglutamate at the N-terminus of
glutaminyl peptides was a spontaneous reaction (3). Since this intramolecular reaction occurs
only very slowly under physiological conditions, glutaminyl cyclase (QC; EC 2.3.2.5) was
postulated and subsequently identified to be the catalyst responsible for the transformation of
N-terminal glutamine residues during posttranslational maturation of peptides in the secretory
pathway of mammals and plants (Scheme 1) (4-7).
The first QC was isolated by Messer from the latex of the tropical plant Carica papaya in
1963 (4). Later, in 1987, a corresponding enzymatic activity was discovered in animal
pituitary. This mammalian QC was shown to convert the Gln1-precursors of TRH and GnRH
into the appropriate mature pGlu1-peptides (5;7). In addition, QC was co-localized to the
secretory pathway of the bovine pituitary together with its putative products of catalysis,
supporting its processing role in peptide hormone biosynthesis (8).
Coincidently, in several neurodegenerative disorders pyroglutamate-containing peptides are
thought to contribute to the pathogenesis by enhancing the proteolytic stability and
neurotoxicity of hydrophobic, plaque-forming peptides (9). The most prominent severe
dementia, Alzheimer’s disease (AD), is characterized by abnormal accumulation of
extracellular amyloidotic plaques closely associated with dystrophic neurons, reactive
astrocytes and microglia (10-16). Amyloid-β peptides are generated by proteolytic processing
of the β-amyloid precursor protein (APP), which is cleaved N-terminally by β-secretase
(BACE) and C-terminally by γ-secretase in a subsequent step (17-19) (scheme 2). It should be
mentioned that there is much controversy concerning the ultimate involvement of the γ-
secretase activity of the presenilins in the formation of Aβ(1-40/42) and their utility as targets
in AD-therapy (20;21).
Within the widely heterogeneous N-terminus of the amyloid peptide found in senile plaques
exists a dominant fraction of Aβ-peptides containing an amino terminal aspartate residue such
as
Asp1-Aβ(1-40)
and
Asp1-Aβ(1-42).
These
full-length
Aβ-peptides
are
found
predominantly in the plaque periphery. In contrast, a second dominant fraction of Aβpeptides, those with an N-terminal pyroglutamine, e.g. pGlu3-Aβ(3-40/42) and pGlu11-Aβ(1140/42), can be found preferentially in the core of senile plaques. These shortened peptides are
reported to be more neurotoxic in vitro and to aggregate more rapidly than the full-length
isoforms (9;22). N-terminally truncated peptides have been shown to be overproduced in early
onset familial AD (FAD) subjects (23), to be dominant in diffuse plaques in the brains of
patients with Down´s syndrome (DS) and AD (24), and to appear early and increase with age
in Down’s syndrome brains (25-28). Further, their quantity has been shown to correlate with
disease severity (25).
Among all prominent Aβ peptides, the isoforms containing pyroglutamate at position 3, such
as pGlu3-Aβ(3-40/42), represent the most abundant of the N-terminally truncated peptide
species (~ 50 % of total Aβ protein), particularly in the core of senile plaques (29-31). The
accumulation of pGlu-Aβ peptides is likely due to the structural modification that enhances
aggregation and confers resistance to most aminopeptidases (23;32). This evidence provides
clues for a pivotal role of pGlu-Aβ peptides in AD pathogenesis. Cyclization, which yields
the pyroglutamate form of an amyloidogenic peptide (with an uncharged N-terminus), may
contribute to the overall hydrophobicity of the structure (9).
There are four potential pathways, which could lead to such neurotoxic pGlu-compounds:
(i)
spontaneous cyclization of N-terminal glutamate residues following exposure by
BACE and/or aminopeptidases
(ii)
Glu to Gln mutations and/or posttranslational esterification or amidation of Aβglutamates buried within the APP-chain and subsequent spontaneous cyclization
of the glutamines after N-terminal exposure by BACE and/or aminopeptidase
processing
(iii)
enzymatic cyclization of N-terminally uncovered glutamate
(iv)
Glu to Gln mutations and/or posttranslational amidation of the Aβ-glutamates
buried within the APP-chain and enzymatic cyclization after N-terminal exposure
by BACE and/or aminopeptidase processing.
So far, there is no experimental evidence that is supportive of pathways (i) and (ii). The
enzymatic conversion of Glu1-peptides into pGlu1-peptides by an unknown glutamyl cyclase
(EC) corresponding to pathway (iii) was recently proposed (30). However, to date, no such
enzyme activity has been identified, capable of cyclizing Glu1-peptides which are Nterminally protonated and possess a negatively charged Glu1 γ-carboxylate moiety. Hence, the
remaining postulated path to N-terminal pGlu formation (iv) may involve glutaminyl cyclase
activity as speculated previously (33). However, QC-activity against Gln1-substrates is
dramatically reduced below pH 7.0 (Schilling et. al., 2003a1). Interestingly, Glu1-conversion
has been reported to occur at acidic reaction conditions (24-30).
In order to prove whether QC is able to recognize and to turnover amyloid-β derived peptides
under mild acidic conditions, we synthesized and investigated Gln3-Aβ(1-11)a, Aβ(3-11)a,
Gln3-Aβ(3-11)a, Aβ(3-21)a, Gln3-Aβ(3-21)a and Gln3-Aβ(3-40) as potential substrates of the
1
Schilling, S., Manhart, T., Hoffmann, T., Ludwig, H.-H., Wasternack, C. and Demuth, H.-U., Substrate
Specificity of Glutaminyl Cyclases from Plants and Animals, Biol. Chem. (2003) 384, in press
enzyme. These sequences were designed according to the sequences of naturally occuring Nterminally and C-terminally truncated Glu3-Aβ peptides and Gln3-Aβ peptides which could
occur due to posttranslational Glu-amidation.
Another objective of the study was to compare the fate of the amyloid-β derived peptides
using either porcine pituitary homogenate as source of native QC, purified recombinant
human QC alone, or QC in combination with an aminopeptidase, under the rationale that
aminopeptidase cleavage of full length Aβ peptide is a pre-requisite to QC cyclization of
Glu/Gln at position three from the N-terminus. Finally, it was of interest whether human QCactivity processing of the above amyloid-β peptides can be suppressed by recently
characterized QC-inhibitors (Schilling et al., 2003b2).
Experimental Procedures
QC Isolation
Human QC was expressed in P. pastoris or E. coli and purified as described (34). Papaya QC
was purified from papaya latex essentially as described elsewhere (35) with the addition of a
third chromatography step on an UNO S (BioRad) column.
Oligopeptide synthesis
Amyloid-β peptide fragments Gln3-Aβ(1-11)a, Aβ(3-11)a, Gln3-Aβ(3-11)a, Aβ(3-21)a and
Gln3-Aβ(3-21)a were synthesized as C-terminal amides both semi-automatically in a 0.5
mmol scale on a peptide synthesizer (Labortec SP650, Bachem) as previously described (34)
or using an automated Symphony peptide synthesizer (Rainin Instrument Co.) in a 50µmol
scale. Amyloid-β peptide Gln3-Aβ(3-40) was synthesized in 25 µmol scale on Fmoc-Val2
Schilling, S., Niestroj, A.J., Rahfeld, J.-U., Hoffmann, T., Wermann, M., Zunkel, K., Wasternack, C. and
Demuth, H.-U., Identification of Human Glutaminyl Cyclase as a Metalloenzyme: Potent Inhibition by
NovaSyn®TGA resin (0.15 mmol/g) or on the NovaSyn®TGR resin (0.23mmol/g) using the
automated Symphony peptide synthesizer. For all peptide couplings modified Fmoc-protocols
of solid-phase peptide synthesis were employed using 2-(1H-Benzotriazole-1-yl)-1,1,3,3,tetramethyluronium tetrafluoroborate (TBTU; Novabiochem)/N-methyl-morpholine; (NMM;
Merck) as coupling regents. The coupling reaction was carried out by using 5 eq of Fmocamino acid, 5 eq TBTU and 10 eq NMM employing double coupling (twice 30 min and twice
1h from the 21st coupling onwards). Deprotection was carried out by using 20% piperidine in
DMF (twice 5 min, then from the 21st coupling step onwards, once 5 min and once 12 min).
After cleavage from the resin by a trifluoroacetic acid (TFA; Merck) containing cocktail, the
crude peptides were purified by preparative HPLC with acid free solvents in order to avoid
further cyclization of the N-terminal glutamine. Preparative HPLC was performed with a
gradient of acetonitrile in water (20 % to 65 % acetonitrile over 40 min) on a 250-10 Luna
RP18, WP300 column (MERCK). To confirm peptide purity and identity analytical HPLC
and ESI-MS were performed.
Assays of QC
QC activity was evaluated fluorometrically using Gln-βNA at 30 °C, essentially as described
(35). Glu-βNA was employed for observation of glutamate cyclization catalyzed by papaya
QC. Spontaneous cyclization of Glu-βNA and Gln-βNA (2 mM) was investigated for 30 days
in 20 mM MES-buffer, pH 6.0, 30°C. Samples were removed for determination of pGlucontent, diluted 10-fold and fluorometrically analyzed (35).
For the pH-dependence studies under first-order rate-law conditions (i.e. substrate
concentrations far below KM-values), a buffer was prepared consisting of 0.05 M acetic acid,
0.05 M pyrophosphoric acid and 0.05 M Tricine. The pH-value was adjusted by addition of
NaOH. In order to compensate for differences in ionic strength between buffers, constant
Imidazole Derivatives and Heterocyclic Chelators. J. Biol. Chem. (2003) 278, in press
conductivity was maintained by addition of NaCl to the different buffers prepared at different
pH-values. The spectrophotometric assay of QC was applied as described elsewhere
(Schilling et al., 2003a). Enzyme kinetic data was analyzed using Grafit software (version
5.0.4. for windows, Erithacus software Ltd., Horley, UK).
MALDI-TOF mass spectrometry
Matrix-assisted laser desorption/ionization mass spectrometry was carried out using a
Hewlett-Packard G2025 LD-TOF System. Enzymatic reactions using the Gln-peptides were
performed in samples of 100 µl consisting of QC (0.01- 1 U) and 0.5 mM substrate in 0.04 M
Tris/HCl, pH 8.0, at 30 °C or at pH- and buffer conditions described further in the text. At the
times indicated, samples were removed, diluted with matrix and analyzed as described
previously (34).
For long-term testing of Glu1-cyclization, Aβ-derived peptides were incubated in 100µl 0.1 M
sodium acetate buffer, pH 5.2 or 0.1 M Bis-Tris buffer, pH 6.5 at 30°C. Peptides were applied
in 0.5 mM [Aβ(3-11)a and other synthetic peptides] or 0.15 mM [Aβ(3-21)a] concentrations,
and 0.2 U QC was added all 24 hours. In the case of Aβ(3-21)a, the assays contained 1 %
DMSO. At the times indicated in the Figures 1-3, samples were removed from the assay tube,
peptides were extracted using ZipTips (Millipore) according to the manufacturer`s
recommendations, mixed with matrix solution (1:1 v/v) and subjected to mass spectrometry.
Negative controls did either contain no QC or heat deactivated enzyme. For the inhibitor
studies the sample composition was the same as described above, with exception of the
inhibitory compound added (5 mM benzimidazole or 2 mM 1,10-phenanthroline).
Results
Synthesis of Aβ peptides
The synthesis of Aβ(1-40/42) peptides is known to be difficult due to excessive
hydrophobicity in the C-terminal region which leads to aggregation or secondary structure
formation during stepwise solid-phase peptide synthesis. Several attempts have been made to
overcome these difficulties for instance by using stronger coupling reagents such as HATU or
more efficient base for deprotection like DBU (36;37). To increase coupling yields during
synthesis of Gln3-Aβ(3-40) we used a low pre-loaded resin (Fmoc-Val-NovaSyn®TGA, 0.15
mmol/g). Furthermore, we introduced a pseudoproline unit (Fmoc-Gly-Ser(ΨMe,Mepro)-OH
instead of Gly25-Ser26 to disrupt aggregation of the peptide chain (38). The introduction of this
pseudoproline unit resulted in a significant improvement in the purity and the yield of the
crude amyloid peptide.
Turnover of Gln3-Aβ peptides 3-11a, 3-21a and 3-40 by recombinant human QC
In previous work we characterized more than 30 Gln1- and Glu1-peptides of different chain
lengths as potential substrates of glutaminyl cyclase (Schilling et. al., 2003a). While H-Gln-
βNA, H-Gln-Phe-Lys-Arg-Leu-NH2 and even H-β-homoGln-Phe-Lys-Arg-Leu-Ala-NH2
were recognized and N-terminally cyclized by human QC, compounds such as H-Glu(OMe)Phe-Lys-Arg-Leu-Ala-NH2, H-Glu-βNA or H-Glu-Phe-Lys-Arg-Leu-Ala-NH2 neither served
as substrates nor as inhibitors of the enzyme under the basic pH-conditions applied.
However,
Glu(OMe)-Phe-Lys-Arg-Leu-Ala-NH2
demonstrated
a
tendency
towards
spontaneous formation of pGlu-Phe-Lys-Arg-Leu-Ala-NH2. Consistent with these findings,
Glu(OMe)3-Aβ(3-11)a was also found to cyclize spontaneously (data not shown).
All Gln3-Aβ derived peptides tested were efficiently converted by human QC into the
corresponding pyroglutamyl forms (Table 1). Due to the poor solubility of Gln3-Aβ(3-21)a
and Gln3-Aβ(3-40) in aqueous solution, the determinations were carried out in the presence of
1% DMSO. The higher solubility of Gln3-Aβ(3-11)a, however, enabled kinetic analysis both
in the presence and absence of DMSO (Table 1). Taken together, the investigation of the Aβ
peptides as QC-substrates with chain-length of 8, 18 and 37 amino acids (see Table 1)
confirmed our previous observation that human QC-activity increases with the length of its
substrates. Accordingly, Gln1-gastrin, Gln1-neurotensin, Gln1-GnRH are among the best QCsubstrates when taking the specificity constants into account (Schilling et. al., 2003a).
Similarly, Gln3-Aβ(3-40) and glucagon, the largest QC-substrates investigated thus far,
exhibited high second order rate constants (449 mM-1s-1 and 526 mM-1s-1 respectively) even in
the presence of 1% DMSO (Table 1).
Interestingly, the kinetic parameters for the conversion of the investigated amyloid peptides
did not change dramatically with increasing size, suggesting only moderate effects of the Cterminal part of Aβ on QC catalysis. Therefore, due to better solubility and experimental
handling, the further investigations concerning N-terminal aminopeptidase processing of these
peptides were performed using the smaller fragments of Aβ, Gln3-Aβ(1-11)a, Gln3-Aβ(3-11)a
and Aβ(3-11)a.
Processing of Gln3-Aβ(1-11)a by purified DP IV and QC present in porcine pituitary
homogenate
After β-secretase post-methionine cleavage at position 670 of APP, further N-terminal
degradation of the resulting Aβ peptide(s) occurs until the decomposition is halted by
formation of N-terminal pGlu in vivo (22-28). Such concerted posttranslational processing
mediated by aminopeptidases, dipeptidyl peptidase IV (DP IV) and glutaminyl cyclase has
been already proposed for the formation of mature neuropeptides (39). Since full length Aβ(142) starts with the dipeptide Asp-Ala (a DP IV-recognition sequence) before the Glu in
position 3, we investigated whether purified DP IV or aminopeptidases of porcine pituitary
homogenate were able to remove this sequence from our sample peptides.
Incubation of the model peptides Gln3-Aβ(1-11)a with DP IV and Gln3-Aβ(3-11)a with
porcine pituitary homogenate resulted in the formation of Gln3-Aβ(3-11)a and pGlu3-Aβ(311)a, respectively (see Figure 1A, 1C).
When the reaction was conducted in the presence of the DP IV-inhibitor Val-Pyrr, no further
turnover of Gln3-Aβ(1-11)a was observed (Figure 1B). Similarly, in the presence of the QCinhibitor 1,10-phenanthroline, no final N-terminal pGlu-formation occurred, resulting in
build-up of Gln3-Aβ(3-11) as the final reaction product (Figure 1D).
When Gln3-Aβ(1-11)a was incubated with porcine pituitary homogenate in the absence of
both inhibitors, a slow stepwise removal of both amino acids by brain aminopeptidases and
final cyclization to pGlu3-Aβ(3-11)a by porcine pituitary QC and by DP IV-containing
pituitary homogenate takes place (Figure 2A).
In the presence of the QC-inhibitor however, only the slow processing by brain
aminopeptidases can be observed, yielding the intermediates Gln3-Aβ(2-11)a and Gln3-Aβ(311)a (Figure 2B). By incubating Gln3-Aβ(1-11)a with DP IV-containing porcine pituitary
homogenate in the presence of the DP IV-inhibitor Val-Pyrr only the formation of Gln3-Aβ(211)a was detectable (Figure 2C).
Turnover of Aβ(3-11)a and Aβ(3-21)a by recombinant human QC
The incubation of Aβ(3-11)a and Aβ(3-21)a in the presence of QC revealed that in contrast to
previous work, glutamate-containing peptides can also serve as QC-substrates (Figures 3C
and D). The QC-catalyzed formation of pGlu3-Aβ(3-11)a and pGlu3-Aβ(3-21)a was
investigated at pH 5.2 and 6.5, respectively. If the QC-inhibitor benzimidazole was added to
the solution before starting the assay by the addition of QC, substrate conversion resulting in
pGlu3-Aβ(3-11)a or pGlu3-Aβ(3-21)a was suppressed (Figures 3E and F). If QC was boiled
before addition, formation of the pGlu-peptides was negligible (Figures 3A and B).
pH-dependency of the papaya QC-catalyzed cyclization of Gln-βNA and Glu-βNA
The plant QC from C. papaya, an analogous but non-homologous enzyme of the mammalian
QCs, has been shown to possess a very similar substrate specificity pattern to human QC
(Schilling et al., 2003a). Accordingly, plant QC cyclized Aβ(3-11)a at pH 5.2 (not shown). In
contrast to human QC, however, we also observed the conversion of the short fluorogenic
substrate Glu-βNA, which enabled us to apply a continuous coupled fluorometric assay (35).
Since a different impact of substrate protonation on QC-catalysis was expected, higher
amounts of QC were applied in the model reactions (34;35). Papaya QC converted Glu-βNA
in a concentration range up to 2 mM (which was limited by substrate solubility) in accordance
with Michaelis-Menten kinetics (Figure 4). Inspection of turnover versus substrate
concentration diagrams for the conversion of Glu-βNA between pH 6.1 and 8.5 revealed that
both KM and kcat changed in a pH-dependent manner (Figure 4). This is in contrast to the
previously described QC-catalyzed glutamine cyclization, for which only changes in KM were
observed over the given pH range (40).
Subsequently, to study the impact of the proton concentration during Glu- and Gln-cyclization
by QC, we investigated the pH-dependence of cyclization of Glu-βNA and Gln-βNA under
first-order rate-law conditions (i.e. substrate concentrations far below KM-values) (Figure 5).
As expected the cyclization of glutamine has a pH-optimum at pH 8.0, in contrast to the
cyclization of glutamic acid which showed a pH-optimum of pH 6.0. While the specificity
constants at the respective pH-optima differ approximately 80,000-fold, the ratio of QC
versus EC activity around pH 6.0, is only about 8,000.
The non-enzymatic pGlu-formation from Gln-βNA investigated at pH 6.0, was followed for 4
weeks and revealed a first-order rate constant of 1.2*10-7 s-1. However, during the same time
period, no pGlu-βNA was formed from Glu-βNA enabling estimation of a limiting rate
constant for turnover of 1.0*10-9 s-1.
Discussion
Since spontaneous formation of pGlu3-Aβ(3-40) occurs neither in vitro nor in vivo, only
enzymatic cyclization of Glu-Aβ by a putative glutamyl cyclase (EC) or of Gln-Aβ peptides
by the known glutaminyl cyclase (QC) is conceivable. Here, we have shown that papaya and
human QC catalyze both glutaminyl and glutamyl cyclization (Schemes 1 and 3, Figures 1-5).
The subcellular localization of these reactions remains unclear. Neurons can maintain in the
cytosol pH values between 5.5 and 7.2 and cytosolic proteins of nerve cell enzymes are
optimized for function at acidic pHs (41-43). Hence, under mildly acidic conditions preferred
QC-catalyzed EC-reactions as observed in our study can occur in the cytosol.
However, Glu-Aβ peptides have been found to be preferentially generated by β-secretase
processing in the endopasmatic reticulum as Aβ(1-40/42) and in the trans-golgi network as
Aβ(11-40/42) most likely within secretory vesicles (44;45). Interestingly, QC is also localized
in the secretory pathway (4-7) and by coincidence the major pGlu-Aβ peptides pGlu3-Aβ(340/42) and pGlu11-Aβ(11-40/42) have been found in senile plaques of aged and Down
syndrome brains (46).
The primary physiological function of QC is likely terminal hormone processing (maturation)
in endocrine cells by glutamine cyclization prior to, or during the hormone secretion process.
Such secretory vesicles are known to be acidic in pH. Thus, an auxillary/additional function
of the enzyme in the narrow pH-range from 5.0 to 7.0 could be its newly discovered glutamyl
cyclase activity (Scheme 3) transforming Glu-Aβ peptides. However, due to the relatively
inefficient rate of Glu-cyclization compared to Gln-conversion, it is questionable whether the
glutamyl cyclization plays a significant physiological role. In the etiology of
neurodegenerative disorders, however, glutamyl cyclization may be of relevance providing
that accumulation of peptidase resistant substrate, and appropriate QC-concentration and
compartment acidity, coincide.
Investigating the pH-dependency of this enzymatic reaction, we found that the unprotonated
N-terminus was essential for the cyclization of Gln1-peptides and accordingly that the pKavalue of the substrate was identical to the pKa-value for QC-catalysis (see Figure 5 and
Schilling et. al., 2003a). These results support the view that QC stabilizes the intramolecular
nucleophilic attack of the unprotonated α-amino moiety on the γ-carbonyl carbon
electrophilically activated by amidation (Scheme 1).
In contrast to the monovalent charge present on N-terminal glutamine containing peptides, the
N-terminal Glu-residue in Glu-containing peptides is predominantly bivalently charged
around neutral pH. Glutamate exhibits pKa-values of about 4.2 and 7.5 for the γ-carboxylic
and for the α-amino moiety, respectively. I.e. at neutral pH and above, although the α-amino
nitrogen is partially or fully unprotonated and nucleophilic, the γ-carboxylic group is
unprotonated, and so exercising no electrophilic carbonyl activity. Hence, intramolecular
cyclization is impossible.
However, in the pH-range of about 5.2-6.5, between their respective pKa-values, the two
functional groups are present both partially non-ionized, in concentrations of about 1-10% (NH2) or 10-1% (-COOH) of total N-terminal Glu-containing peptide. As a result, over a
mildly acidic pH-range species of N-terminal Glu-peptides are present which carry both
groups uncharged, and, therefore, it is possible that QC could stabilize the intermediate of
intramolecular cyclization to pGlu-peptide. I.e. if the γ-carboxylic group is protonated, the
carbonyl carbon is electrophilic enough to allow nucleophilic attack by the unprotonated αamino group. At this pH the hydroxyl ion functions as a leaving group (Scheme 3). These
assumptions are corroborated by the pH-dependence data obtained for the QC catalyzed
conversion of Glu-βNA. In contrast to glutamine conversion of Gln-βNA by QC, the pHoptimum of catalysis shifts to the acidic range to around pH 6.0, i.e. the pH-range in which
substrate molecule species are simultaneously abundant carrying a protonated γ-carboxyl and
unprotonated α-amino group. Furthermore, the kinetically determined pKa-value of 7.55
±0.02 is in excellent agreement with that of the α-amino group of Glu-βNA, determined by
titration (7.57 ±0.05).
Physiologically, at pH 6.0 the second-order rate constant (or specificity constant, kcat/KM) of
the QC-catalyzed glutamate cyclization might be in the range of 8,000fold slower than the one
for glutamine cyclization (compare data in Figure 4). However, the non-enzymatic turnover of
both model substrates Glu-βNA and Gln-βNA is negligible, which corroborates with the
observed negligible pGlu-peptide formation in our study. Hence, for the pGlu-formation by
QC an acceleration of at least 108 can be estimated from the ratio of the enzymatic versus
non-enzymatic rate constants (comparing the second-order rate constants for the enzyme
catalysis with the respective nonenzymatic cyclization first-order rate constants the catalytic
proficiency factor is about 109 and 1010 M-1 for the Gln- and the Glu-conversion,
respectively). The conclusion from these data is, that in vivo only an enzymatic path resulting
pGlu-formations seems conceivable.
Since QC is highly abundant in the brain and taking into account the high turnover rate of 0.9
min-1 recently found for the maturation of 30 µM of (Gln-)TRH-like peptide (47), one can
predict a cyclization half-life of about 100 hours for an appropriate glutamate-substrate,
providing similar reaction conditions. Moreover, given compartmentalization and localization
of brain QC/EC in the secretory pathway, the actual in vivo enzyme and substrate
concentrations and reaction conditions might be even more favorable for the enzymatic
cyclization in the intact cell. In addition, if N-terminal Glu is transformed to Gln a much more
rapid pGlu-formation mediated by QC could be expected. In vitro, we were able to suppress
both reactions by applying inhibitors of QC/EC-activity (Figures 2 and 3).
Whether N-terminal processing of the tested Aβ-derived peptides takes place in vivo by a
combination of aminopeptidase, dipeptidyl peptidase and glutaminyl/glutamyl cyclase activity
and whether its selective blockage as achieved in our experiments is of physiological
relevance begs further investigation. Previously, such N-terminal processing in an analog
fashion was suggested for the immature neuropeptide Antho-RFamide precursors of
coelenterates (39). Similarly, impaired post-translational proteolytic processing of Aβ
peptides by aminopeptidases has been suggested as a causative event in AD (48). An
imbalance between anabolic and catabolic processes of the APP-biosynthesis and the Aβ
peptide degradation could cause the accumulation of Aβ(1-42) and other Aβ-peptides which
are normally further degraded. According to this hypothesis, such an imbalance would occur
through a reduced aminopeptidase activity which would not sufficiently process the Nterminus of Aβ resulting in a negligible iso-Asp1-Aβ(1-42) formation as compared to a
significant accumulation of pGlu-Aβ peptides (23-29;46).
Recent reviews on potential AD treatment do not propose inhibition of pyroglutamate
formation as potential therapeutic intervention (9;17;18;48). Since pGlu-Aβ peptides appear
highly abundant, because they are even more hydrophobic and neurotoxic than the Nterminally charged amyloid peptides and since pGlu-Aβ formation prevents intracellular
aminopeptidase-mediated disposal of such improperly generated peptides (44;46), inhibition
of brain glutaminyl cyclase QC- and EC-activity could prove a valuable tool to combat the
onset and progression of neurodegenerative disorders. In addition, a combination of the
inhibition of QC/EC and aminopeptidases might be a suitable experimental approach to
further investigate the site of the plaque-formation resulting from pGlu-Aβ peptides in situ
and in vivo.
In summary, our results indicate that human QC/EC, which is highly abundant in the brain, is
a likely catalyst to the formation of the amyloidogenic pGlu-Aβ peptides (from Glu-Aβ and
Gln-Aβ precursors), major constituents of the senile plaques found in AD. In addition to their
contribution towards our basic understanding of peptide processing during hormone
maturation, these findings identify QC/EC as a potential player in plaque formation and thus
as a novel drug target in the treatment of AD.
Acknowledgement
We thank H.-H. Ludwig for excellent technical assistance, M. Wermann for providing a DP
IV-sample and we are grateful to Dr. J.A. Pospisilik for critical comments on the manuscript.
References
1. Awade, A. C., Cleuziat, P., Gonzales, T., and Robert-Baudouy, J. (1994) Proteins 20,
34-51.
2. Garavelli, J. S. (2000) Nucleic Acids Res 28, 209-211.
3. Richter, K., Kawashima, E., Egger, R., and Kreil, G. (1984) EMBO J 3, 617-621.
4. Messer, M. (1963) Nature 4874, 1299.
5. Busby, W. H., Quackenbush, G. E., Humm, J., Youngblood, W. W., and Kizer, J. S.
(1987) J Biol Chem 262, 8532-8536.
6. Dahl, S. W., Slaughter, C., Lauritzen, C., Bateman, R. C., Connerton, I., and Pedersen,
J. (2000) Protein Express Purif 20, 27-36.
7. Fischer, W. H. and Spiess, J. (1987) Proc Natl Acad Sci U S A 84, 3628-3632.
8. Bockers, T. M., Kreutz, M. R., and Pohl, T. (1995) J Neuroendocrinol 7, 445-453.
9. Saido, T. C. (2002) Med Hypotheses 54, 427-429.
10. Terry, R. D. and Katzman, R. (1983) Ann Neurol 14, 497-506.
11. Glenner, G. G. and Wong, C. W. (1984) Biochem Biophys Res Commun 120, 885-890.
12. Itagaki, S., McGeer, P. L., Akiyama, H., Zhu, S., and Selkoe, D. (1989) J
Neuroimmunol 24, 173-182.
13. Funato, H., Yoshimura, M., Kusui, K., Tamaoka, A., Ishikawa, K., Ohkoshi, N.,
Namekata, K., Okeda, R., and Ihara, Y. (1998) Am J Pathol 152, 1633-1640.
14. Selkoe, D. J. (2001) Physiol Rev 81, 741-766.
15. Borchelt, D. R. (1998) Lab Anim Sci 48, 604-610.
16. Mann, D. M. and Iwatsubo, T. (1996) Neurodegeneration 5, 115-120.
17. Citron, M. (2002) Nat Neurosci 5 Suppl: 1055-1057.
18. Wolfe, M. S. (2002) Nat Rev Drug Discov 1, 859-866.
19. Kimberly, W. T., Esler, W. P., Ye, W., Ostaszewski, B. L., Gao, J., Diehl, T., Selkoe,
D. J., and Wolfe, M. S. (2003) Biochemistry 42, 137-144.
20. Thinakaran, G. (1999) J Clin Invest 104, 1321-1327.
21. Wilson, C. A., Doms, R. W., Zheng, H., and Lee, V. M. (2002) Nat Neurosci 5, 849855.
22. Pike, C. J., Overman, M. J., and Cotman, C. W. (1995) J Biol Chem 270, 2389523898.
23. Saido, T. C., Iwatsubo, T., Mann, D. M., Shimada, H., Ihara, Y., and Kawashima, S.
(1995) Neuron 14, 457-466.
24. Iwatsubo, T., Saido, T. C., Mann, D. M., Lee, V. M., and Trojanowski, J. Q. (1996)
Am J Pathol 149, 1823-1830.
25. Russo, C., Saido, T. C., DeBusk, L. M., Tabaton, M., Gambetti, P., and Teller, J. K.
(1977) FEBS Lett 409, 411-416.
26. Russo, C., Salis, S., Dolcini, V., Venezia, V., Song, X. H., Teller, J. K., and Schettini,
G. (2001) Neurobiol Dis 8, 173-180.
27. Tekirian, T. L., Saido, T. C., Markesbery, W. R., Russell, M. J., Wekstein, D. R.,
Patel, E., and Geddes, J. W. (1998) J Neuropathol Exp Neurol. 57, 76-94.
28. Russo, C., Violani, E., Salis, S., Venezia, V., Dolcini, V., Damonte, G., Benatti, U.,
DArrigo, C., Patrone, E., Carlo, P., and Schettini, G. (2002) J Neurochem 82, 14801489.
29. Hosoda, R., Saido, T. C., Otvos, L., Jr., Arai, T., Mann, D. M., Lee, V. M.,
Trojanowski, J. Q., and Iwatsubo, T. (1998) J Neuropathol Exp Neurol 57, 1089-1095.
30. Garden, R. W., Moroz, T. P., Gleeson, J. M., Floyd, P. D., Li, L. J., Rubakhin, S. S.,
and Sweedler, J. V. (1999) J Neurochem 72, 676-681.
31. Harigaya, Y., Saido, T. C., Eckman, C. B., Prada, C. M., Shoji, M., and Younkin, S.
G. (2000) Biochem Biophys Res Commun 276, 422-427.
32. Tekirian, T. L., Yang, A. Y., Glabe, C., and Geddes, J. W. (1999) J Neurochem 73,
1584-1589.
33. Shirotani, K., Tsubuki, S., Lee, H. J., Maruyama, K., and Saido, T. C. (2002) Neurosci
Lett 327, 25-28.
34. Schilling, S., Hoffmann, T., Rosche, F., Manhart, S., Wasternack, C., and Demuth, H.U. (2002) Biochemistry 41, 10849-10857.
35. Schilling, S., Hoffmann, T., Wermann, M., Heiser, U., Wasternack, C., and Demuth,
H.-U. (2002) Anal Biochem 303, 49-56.
36. Tickler, A. K., Barrow, C. J., and Wade, J. D. (2001) J Pept Sci 7, 488-494.
37. Fukuda, H., Shimizu, T., Nakajima, M., Mori, H., and Shirasawa, T. (1999) Bioorg
Med Chem Lett 9, 953-956.
38. Mutter, M., Nefzi, A., Sato, T., Sun, X., Wahl, F., and Wohr, T. (1995) Pept Res 8,
145-153.
39. Schmutzler, C., Darmer, D., Diekhoff, D., and Grimmelikhuijzen, C. J. (1992) J Biol
Chem 267, 22534-22541.
40. Gololobov, M. Y., Song, I., Wang, W., and Bateman, R. C. (1994) Arch Biochem
Biophys 309, 300-307.
41. Nachshen, D. A. and Drapeau, P. (1988) J Gen Physiol 91, 289-303.
42. Drapeau, P. and Nachshen, D. A. (1988) J Gen Physiol 91, 305-315.
43. Ruscak, M., Orlicky, J., Zubor, V., and Hager, H. (1982) J Neurochem 39, 210-216.
44. Huse, J. T., Liu, K., Pijak, D. S., Carlin, D., Lee, V. M., and Doms, R. W. (2002) J
Biol Chem 277, 16278-16284.
45. Haass, C., Lemere, C. A., Capell, A., Citron, M., Seubert, P., Schenk, D., Lannfelt, L.,
and Selkoe, D. J. (1995) Nat Med 1, 1291-1296.
46. Atwood, C. S., Martins, R. N., Smith, M. A., and Perry, G. (2002) Peptides 23, 13431350.
47. Prokai, L., Prokai-Tatrai, K., Ouyang, X., Kim, H. S., Wu, W. M., Zharikova, A., and
Bodor, N. (1999) J Med Chem 42, 4563-4571.
48. Saido, T. C. (1998) Neurobiol Aging 19, 69-75.
Tables
Table 1: Kinetic parameters for conversion of N-terminally Gln-containing peptides by
recombinant human QC in buffer solution containing 1% DMSO
KM (µM)
kcat (s-1)
kcat/KM (mM-1s-1)
Gln3-Aβ(3-11)a
87 ±3#
55 ±1#
632 ±10#
Gln3-Aβ(3-11)a
155 ±4
41.4 ±0.4
267 ±4
Gln3-Aβ(3-21)a
162 ±12
62 ±3
383 ±10
Gln3-Aβ(3-40)
89 ±10
40 ±2
449 ±28
Glucagon(3-29)
19 ±1
10.0 ±0.2
526 ±17
Peptide
#
Determined in absence of DMSO
Schemes
Scheme 1:
N-terminal cyclization of glutaminyl peptides by QC
peptide
NH
O
O
O
NH2
NH
QC
O
H2N
O
NH3
NH2
QC
Scheme 2:
HN
HN
H2N
peptide
peptide
O
Sequence fragment of human amyloid precursor protein (APP 770)
661 Ile Lys Thr Glu Glu Ile Ser Glu Val Lys Met β Asp Ala Glu Phe
1
2
3
675
4
676 Arg His Asp Ser Gly Tyr Glu Val His His Gln Lys α Leu Val Phe 690
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
691 . Phe Ala Glu Asp Val Gly Ser Asn Lys Gly Ala Ile Ile Gly Leu
20
21
22
23
24
25
26
27
28
29
30
31 32
33
705
34
706 Met Val Gly Gly Val Val Ile Ala γ Thr Val Ile Val Ile Thr Leu
35
•
•
•
•
Scheme 3:
H3N
36
37
38
39
40
41
Full length Aβ peptide sequence 1-42 corresponding to amino acids 672-713
of APP in black.
α-, β-, γ-secretase cleavage sites marked bold and underscored
Glu3 and Glu11 of the Aβ peptide marked bold and underscored.
N-terminal aminopeptidase processing site Asp-Ala are in italics.
N-terminal cyclization of uncharged glutamyl peptides by QC (EC)
peptide
peptide
NH
NH
O
H2N
peptide
HN
O
QC/EC
O
NH2
O
O
720
42
O
OH
HO
O
peptide
HN
O
H2O
QC/EC
NH
O
Figure Legends
Figure 1:
A Mass spectra of Gln3-Aβ(1-11)a incubated with DP IV catalyzing the N-
terminal truncation yielding Gln3-Aβ(3-11)a. B Mass spectra of Gln3-Aβ(1-11)a incubated
with DP IV and the DP IV-inhibitor Val-Pyrrolidide (Val-Pyrr) preventing N-terminal
truncation of the peptide. C Mass spectra of Gln3-Aβ(3-11)a incubated with porcine pituitary
homogenate catalyzing the formation of pGlu3-Aβ(3-11)a. D Mass spectra of Gln3-Aβ(3-11)a
incubated with QC and the QC-inhibitor 1,10-phenanthroline preventing the formation of
pGlu3-Aβ(3-11)a.
Figure 2:
A Mass spectra of Gln3-Aβ(1-11)a incubated with DP IV-containing porcine
pituitary homogenate resulting in the formation of pGlu3-Aβ(3-11)a after consecutive
catalysis by both enzymes. B Mass spectra of Gln3-Aβ(1-11)a in the presence of DP IVcontaining porcine pituitary homogenate and the QC-inhibitor 1,10-phenanthroline preventing
pGlu3-Aβ(3-11)a formation. C Mass spectra of Gln3-Aβ(1-11)a in the presence of DP IVcontaining porcine pituitary homogenate and the DP IV-inhibitor Val-Pyr suppressing the
formation of pGlu3-Aβ(3-11)a.
Figure 3:
A and B Mass spectra of Glu3-Aβ(3-11)a and Glu3-Aβ(3-21)a incubated with
recombinant human QC which was boiled for 10 min before use. C and D Mass spectra of
Glu3-Aβ(3-11)a and Glu3-Aβ(3-21)a in the presence of active human QC resulting in the
formation of pGlu3-Aβ(3-11)a and pGlu3-Aβ(3-21)a, respectively. E and F Mass spectra of
Glu3-Aβ(3-11)a and Glu3-Aβ(3-21)a in the presence of active QC and 5 mM Benzimidazole
suppressing the formation of pGlu3-formation.
Figure 4:
Reaction rates of papaya QC- catalyzed Glu-βNA-conversion plotted against
the substrate concentration. The initial rates were measured in 0.1 M pyrophosphate buffer,
pH 6.1 (squares), 0.1 M phosphate buffer, pH 7.5 (circles) and 0.1 M borate buffer, pH 8.5
(triangles). The kinetic parameters were as follows: KM= 1.13 ±0.07 mM, kcat= 1.13 ±0.04
min-1 (pH 6.1); KM= 1.45 ±0.03 mM, kcat= 0.92 ±0.01 min-1 (pH 7.5); KM= 1.76 ±0.06 mM,
kcat= 0.56 ±0.01 min-1 (pH 8.5).
Figure 5:
pH-dependence of the conversion of Gln-βNA (circles) and Glu-βNA
(squares), determined under first-order rate-law conditions (S<<KM). Substrate concentration
were 0.01 mM and 0.25 mM, respectively. For both determinations, a three-component buffer
system was applied consisting of 0.05 M acetic acid, 0.05 M pyrophosphoric acid and 0.05 M
Tricine. All buffers were adjusted to equal conductivity by addition of NaCl, in order to avoid
differences in ionic strength. The data were fitted to equations that account for two
dissociating groups revealing pKa-values of 6.91 ±0.02 and 9.5 ±0.1 for Gln-βNA and 4.6
±0.1 and 7.55 ±0.02 for Glu-βNA. The pKa-values of the respective substrate amino groups,
determined by titration, were 6.97 ±0.01 (Gln-βNA) and 7.57 ±0.05 (Glu-βNA). All
determinations were carried out at 30 °C.
Figure 1
A
C
Gln3 -Aβ(3-11)a
pGlu3 -Aβ(3-11)a
t= 90 min
t= 60 min
t= 30 min
relative intensity
relative Intensity
t= 60 min
t= 30 min
t= 10 min
t= 10 min
Gln3 -Aβ(1-11)a
800
1000
1200
t= 5 min
Gln3 -Aβ(3-11)a
t= 1 min
1400
800
1600
1000
t= 1 min
1200
Gln3 -Aβ(3-11)a
Gln3 -Aβ(1-11)a
t= 60 min
relative intensity
t= 45 min
t= 10 min
relative intensity
t= 90 min
t= 30 min
t= 10 min
t= 1 min
t= 1 min
1000
1200
1600
D
B
800
1400
mass/charge ratio (m/z)
mass/charge ratio (m/z)
1400
mass/charge ratio (m/z)
1600
800
1000
1200
mass/charge ratio (m/z)
1400
1600
Figure 2
A
pGlu3 -Aβ(3-11) a
t= 60 min
relative intensity
t= 30 min
t= 10 min
Gln3 -Aβ(1-11)a
t= 0 min
1000
B
1200
1400
1600
mass/charge ratio (m/z)
Gln3 -Aβ(3-11)a
t=90 min
relative intensity
Gln3-Aβ(2-11)a t=60 min
t= 30 min
t=10 min
Gln3 -Aβ(1-11)a
1000
1200
t=0 min
1400
1600
mass/charge ratio (m/z)
C
pGlu3-Aβ(3-11) a
t= 90 min
Gln3 -Aβ(2-11)a
relative intensity
t= 60 min
t= 30 min
Gln3-Aβ(1-11)a
t= 0 min
1000
1200
1400
mass/charge ratio (m/z)
1600
Figure 3
B
A
t= 72 h
relative intensity
relative intensity
t= 72 h
t= 48 h
t= 24 h
t= 48 h
t= 23 h
Glu3-Aβ(3-21)a
Glu3-Aβ(3-11)a
t= 0 h
t= 0 h
800
1000
1200
1400
2000
1600
D
pGlu3-Aβ(3-11)a
pGlu3-Aβ(3-21)a
relative intensity
t= 24 h
t= 48 h
t= 23 h
t= 9 h
Glu3-Aβ(3-11)a
Glu3-Aβ(3-21)a
t= 0 h
1000
1200
1400
1600
2000
mass/charge ratio (m/z)
E
2400
2600
F
relative intensity
t= 48 h
relative intensity
t= 48 h
t= 24 h
t= 23 h
Glu3-Aβ(3-11)a
Glu3-Aβ(3-21)a
t= 0 h
1200
mass/charge ratio (m/z)
1400
t= 72 h
pGlu3-Aβ(3-21)a
t= 72 h
1000
2200
t= 0 h
mass/charge ratio (m/z)
pGlu3-Aβ(3-11)a
800
2600
t= 72 h
t= 72 h
t= 48 h
800
2400
relative intensity
C
2200
mass/charge ratio (m/z)
mass/charge ratio (m/z)
t= 0 h
1600
2000
2200
2400
mass/charge ratio (m/z)
2600
Figure 4
β NA formation (µM/min)
0.6
pH 6.1
pH 7.5
0.4
pH 8.5
0.2
0
0
0.5
1
1.5
[Glu-β NA] (mM)
2
Figure 5
1000
0,012
0,01
800
0,008
600
0,006
400
0,004
200
0,002
0
0
5
6
7
pH
8
kcat/KM Gln-ßNA (mM -1s-1)
kcat/KM Glu-ßNA (mM -1s-1)
1200
0,014